{ "object": "list", "data": [ { "id": "fugu-ultra", "name": "Sakana: Fugu Ultra", "short_name": "Fugu Ultra", "description": "Fugu Ultra is the higher-performance model in Sakana AI's Fugu family. Rather than a single monolithic model, Fugu is a learned multi-agent orchestration system: a language model trained to route tasks across a swappable pool of underlying models and to recursively call instances of itself.", "context_length": 1000000, "created": 1782086400, "owned_by": "sakana", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 5.0 }, "output": { "unit": "1M tokens", "price_per_million_t": 30.0 } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": true, "function_call": true, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": null, "artificial_analysis_coding_index": null, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": null, "hle": null, "livecodebench": null, "scicode": null, "math_500": null, "aime": null, "aime_25": null, "ifbench": null, "lcr": null, "terminalbench_hard": null, "tau2": null } }, { "id": "north-mini-code", "name": "Cohere: North Mini Code 1.0", "short_name": "North Mini Code 1.0", "description": "North Mini Code is an open weights research release of a 30B-A3B parameter model optimized for code generation, agentic software engineering, and terminal tasks.", "context_length": 256000, "created": 1781740800, "owned_by": "cohere", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.2 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.8 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0.27 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": false, "function_call": true, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": null, "artificial_analysis_coding_index": null, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": null, "hle": null, "livecodebench": null, "scicode": null, "math_500": null, "aime": null, "aime_25": null, "ifbench": null, "lcr": null, "terminalbench_hard": null, "tau2": null } }, { "id": "glm-5.2", "name": "Z.ai: GLM 5.2", "short_name": "GLM 5.2", "description": "GLM-5.2 is Z.AI's flagship model for long-horizon autonomous coding and engineering workflows. It is built to plan, execute, iterate, and optimize complex development tasks over extended runs", "context_length": 1000000, "created": 1781481600, "owned_by": "z-ai", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 1.4 }, "output": { "unit": "1M tokens", "price_per_million_t": 4.4 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0.27 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": false, "function_call": true, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": null, "artificial_analysis_coding_index": null, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": null, "hle": null, "livecodebench": null, "scicode": null, "math_500": null, "aime": null, "aime_25": null, "ifbench": null, "lcr": null, "terminalbench_hard": null, "tau2": null } }, { "id": "kimi-k2.7-code", "name": "MoonshotAI: Kimi K2.7 Code", "short_name": "Kimi K2.7 Code", "description": "Kimi K2.7 Code is Moonshot AI's coding-focused agentic model built for long-horizon software engineering workflows. It supports native image input, tool calling, and forced thinking mode.", "context_length": 262144, "created": 1781222400, "owned_by": "moonshot-ai", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.95 }, "output": { "unit": "1M tokens", "price_per_million_t": 4 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0.2 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": true, "function_call": true, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": 53.8, "artificial_analysis_coding_index": 45.5, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": 0.866, "hle": 0.338, "livecodebench": null, "scicode": 0.502, "math_500": null, "aime": null, "aime_25": null, "ifbench": 0.799, "lcr": 0.733, "terminalbench_hard": 0.432, "tau2": 0.942 } }, { "id": "mimo-v2.5-pro-ultraspeed", "name": "Xiaomi: MiMo V2.5 Pro UltraSpeed", "short_name": "MiMo V2.5 Pro UltraSpeed", "description": "MiMo V2.5 Pro UltraSpeed is Xiaomi's speed-focused 1T-parameter MiMo V2.5 Pro mode, built for near-instant coding assistance, real-time chat, live edits, and low-latency agent loops. Xiaomi reports up to roughly 1,000 tokens per second using its TileRT serving stack, FP4 expert quantization, and DFlash speculative decoding.", "context_length": 1000000, "created": 1781049600, "owned_by": "xiaomi", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 1.5 }, "output": { "unit": "1M tokens", "price_per_million_t": 3 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0.12 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": false, "function_call": true, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": 53.8, "artificial_analysis_coding_index": 45.5, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": 0.866, "hle": 0.338, "livecodebench": null, "scicode": 0.502, "math_500": null, "aime": null, "aime_25": null, "ifbench": 0.799, "lcr": 0.733, "terminalbench_hard": 0.432, "tau2": 0.942 } }, { "id": "claude-fable-5", "name": "Anthropic: Claude Fable 5", "short_name": "Claude Fable 5", "description": "Claude Fable 5 is Anthropic's frontier Mythos-class model for demanding engineering work, with standout performance on large codebases and agentic programming. It is built for compressing complex programming projects from months into days", "context_length": 1000000, "created": 1780963200, "owned_by": "anthropic", "available": false, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 10 }, "output": { "unit": "1M tokens", "price_per_million_t": 50 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 1 }, "write": { "5m": { "unit": "1M tokens", "price_per_million_t": 12.5 }, "1h": { "unit": "1M tokens", "price_per_million_t": 20 } } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": true, "function_call": true, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": 64.9, "artificial_analysis_coding_index": 62, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": 0.926, "hle": 0.533, "livecodebench": null, "scicode": 0.602, "math_500": null, "aime": null, "aime_25": null, "ifbench": 0.63469387755102, "lcr": 0.7, "terminalbench_hard": 0.628787878787879, "tau2": 0.985380116959064 }, "discount": { "percentage": 5, "note": "A 5% discount applies to input and output token pricing only." } }, { "id": "nemotron-3-ultra-550b-a55b", "name": "NVIDIA: Nemotron 3 Ultra 550B", "short_name": "Nemotron 3 Ultra 550B", "description": "Nemotron-3 Ultra 550B-A55B is NVIDIA's largest open LLM yet, featuring 550B total parameters with 55B active parameters per token using a Mixture-of-Experts architecture. It's specifically built for long-running agentic AI workflows, supporting up to 1M token context and delivering up to 5x faster inference with 30% lower cost for complex agent tasks.", "context_length": 1000000, "created": 1780531200, "owned_by": "nvidia", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.5 }, "output": { "unit": "1M tokens", "price_per_million_t": 2.5 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0.15 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": false, "function_call": true, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": 47.7, "artificial_analysis_coding_index": 37.6, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": 0.867, "hle": 0.266, "livecodebench": null, "scicode": 0.399, "math_500": null, "aime": null, "aime_25": null, "ifbench": 0.813605442176871, "lcr": 0.67, "terminalbench_hard": 0.363636363636364, "tau2": 0.833333333333333 } }, { "id": "qwen3.7-plus", "name": "Qwen: Qwen 3.7 Plus", "short_name": "Qwen 3.7 Plus", "description": "Qwen3.7 Plus is Alibaba's cost-effective Qwen 3.7 multimodal agent model for coding, tool use, productivity workflows, visual understanding", "context_length": 991808, "created": 1780272000, "owned_by": "qwen", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.42 }, "output": { "unit": "1M tokens", "price_per_million_t": 1.68 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0.04 }, "write": { "5m": { "unit": "1M tokens", "price_per_million_t": 0.5 }, "1h": { "unit": "1M tokens", "price_per_million_t": 0.8 } } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": true, "function_call": true, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": 53.3, "artificial_analysis_coding_index": 46.5, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": 0.9, "hle": 0.334, "livecodebench": null, "scicode": 0.455, "math_500": null, "aime": null, "aime_25": null, "ifbench": 0.779591836734694, "lcr": 0.65, "terminalbench_hard": 0.46969696969697, "tau2": 0.929824561403509 }, "sub": true }, { "id": "minimax-m3", "name": "MiniMax: MiniMax M3", "short_name": "MiniMax M3", "description": "MiniMax M3 is an open-weights frontier model built for coding, agent workflows, and multimodal understanding.", "context_length": 512000, "created": 1780272000, "owned_by": "minimax", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.32 }, "output": { "unit": "1M tokens", "price_per_million_t": 1.26 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0.06 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": true, "function_call": true, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": null, "artificial_analysis_coding_index": null, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": null, "hle": null, "livecodebench": null, "scicode": null, "math_500": null, "aime": null, "aime_25": null, "ifbench": null, "lcr": null, "terminalbench_hard": null, "tau2": null }, "sub": true }, { "id": "mirothinker-1-7-deepresearch-mini", "name": "MiroMindAI: MiroThinker 1.7 Deep Research Mini", "short_name": "MiroThinker 1.7 Deep Research Mini", "description": "MiroMind's 30B deep research agent, tuned for lower-cost research with the same 256k context window and build-in web/tool execution.", "context_length": 256000, "created": 1778457600, "owned_by": "miromindai", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 1.25 }, "output": { "unit": "1M tokens", "price_per_million_t": 10 } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": false, "function_call": false, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": null, "artificial_analysis_coding_index": null, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": null, "hle": null, "livecodebench": null, "scicode": null, "math_500": null, "aime": null, "aime_25": null, "ifbench": null, "lcr": null, "terminalbench_hard": null, "tau2": null } }, { "id": "mirothinker-1-7-deepresearch", "name": "MiroMindAI: MiroThinker 1.7 Deep Research", "short_name": "MiroThinker 1.7 Deep Research", "description": "MiroMind's flagship 235B deep research agent for multi-step research with built-in reasoning and web/tool execution.", "context_length": 262000, "created": 1778457600, "owned_by": "miromindai", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 4 }, "output": { "unit": "1M tokens", "price_per_million_t": 25 } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": false, "function_call": false, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": null, "artificial_analysis_coding_index": null, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": null, "hle": null, "livecodebench": null, "scicode": null, "math_500": null, "aime": null, "aime_25": null, "ifbench": null, "lcr": null, "terminalbench_hard": null, "tau2": null } }, { "id": "step-3.7-flash", "name": "StepFun: Step 3.7 Flash", "short_name": "Step 3.7 Flash", "description": "Step 3.7 Flash is StepFun's high-efficiency multimodal MoE model for agentic coding, long-context work, tool use, and native image/video understanding.", "context_length": 256000, "created": 1780012800, "owned_by": "stepfun", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.2 }, "output": { "unit": "1M tokens", "price_per_million_t": 1.15 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0.04 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": true, "function_call": true, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": null, "artificial_analysis_coding_index": null, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": null, "hle": null, "livecodebench": null, "scicode": null, "math_500": null, "aime": null, "aime_25": null, "ifbench": null, "lcr": null, "terminalbench_hard": null, "tau2": null }, "sub": true }, { "id": "sarvam-30b", "name": "Sarvam: Sarvam 30B", "short_name": "Sarvam 30B", "description": "Sarvam 30B is a 30B parameter chat completion model from Sarvam AI with multilingual support, streaming, tool calling, reasoning controls, and a 64k context window.", "context_length": 64000, "created": 1778544000, "owned_by": "savram", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.03 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.11 } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": false, "function_call": true, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": 12.3, "artificial_analysis_coding_index": 7.9, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": 0.633, "hle": 0.07, "livecodebench": null, "scicode": 0.192, "math_500": null, "aime": null, "aime_25": null, "ifbench": 0.264625850340136, "lcr": 0, "terminalbench_hard": 0.0227272727272727, "tau2": 0.345029239766082 }, "sub": true }, { "id": "sarvam-105b", "name": "Sarvam: Sarvam 105B", "short_name": "Sarvam 105B", "description": "Sarvam 105B is a 105B parameter chat completion model from Sarvam AI with multilingual support, streaming, tool calling, reasoning controls, and a 128k context window.", "context_length": 128000, "created": 1778544000, "owned_by": "savram", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.05 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.18 } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": false, "function_call": true, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": 18.2, "artificial_analysis_coding_index": 9.8, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": 0.738, "hle": 0.101, "livecodebench": null, "scicode": 0.264, "math_500": null, "aime": null, "aime_25": null, "ifbench": 0.343537414965986, "lcr": 0, "terminalbench_hard": 0.0151515151515152, "tau2": 0.467836257309941 }, "sub": true }, { "id": "claude-opus-4.8", "name": "Anthropic: Claude Opus 4.8", "short_name": "Claude Opus 4.8", "description": "Claude Opus 4.8 is Anthropic's most capable generally available model to date, optimized for complex reasoning and long-horizon agentic coding. Released on May 28, 2026, it features a 1M token context window, adaptive thinking controls", "context_length": 1000000, "created": 1779926400, "owned_by": "anthropic", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 4.75 }, "output": { "unit": "1M tokens", "price_per_million_t": 23.75 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0.5 }, "write": { "5m": { "unit": "1M tokens", "price_per_million_t": 6.25 }, "1h": { "unit": "1M tokens", "price_per_million_t": 10 } } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": true, "function_call": true, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": 61.4, "artificial_analysis_coding_index": 56.7, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": 0.92, "hle": 0.457, "livecodebench": null, "scicode": 0.535, "math_500": null, "aime": null, "aime_25": null, "ifbench": 0.622448979591837, "lcr": 0.676666666666667, "terminalbench_hard": 0.583333333333333, "tau2": 0.944444444444445 }, "discount": { "percentage": 5, "note": "A 5% discount applies to input and output token pricing only." } }, { "id": "deepseek-v4-pro-routed", "name": "DeepSeek: DeepSeek V4 Pro (Routed)", "short_name": "DeepSeek V4 Pro (Routed)", "description": "DeepSeek V4 Pro is a large-scale Mixture-of-Experts model with a 1M-token context window, designed for advanced reasoning, coding, long-horizon agent workflows, knowledge, math, and software engineering tasks. Note: This version is routed through DeepSeek infrastructure, making it more cost-efficient, but prompts and responses may be used by DeepSeek to improve their models. If you require stronger privacy guarantees, consider using DeepSeek V4 Pro (deepseek-v4-pro), which is not routed to DeepSeek.", "context_length": 1048576, "created": 1776988800, "owned_by": "deepseek-ai", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.44 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.87 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0.04 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "stop", "temperature", "top_p", "tools", "tool_choice" ], "capabilities": { "vision": false, "function_call": true, "reasoning": true } }, { "id": "qwen3.7-max", "name": "Qwen: Qwen 3.7 Max", "short_name": "Qwen 3.7 Max", "description": "Qwen3.7 Max is Alibaba's latest flagship Qwen model for agentic coding, office automation, and long-running tool workflows.", "context_length": 1000000, "created": 1779376861, "owned_by": "qwen", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 2.38 }, "output": { "unit": "1M tokens", "price_per_million_t": 7.13 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0.25 }, "write": { "5m": { "unit": "1M tokens", "price_per_million_t": 3.13 }, "1h": { "unit": "1M tokens", "price_per_million_t": 5 } } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "stop", "temperature", "top_p", "tool_choice", "tools" ], "capabilities": { "vision": true, "function_call": true, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": null, "artificial_analysis_coding_index": null, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": null, "hle": null, "livecodebench": null, "scicode": null, "math_500": null, "aime": null, "aime_25": null, "ifbench": null, "lcr": null, "terminalbench_hard": null, "tau2": null } }, { "id": "gemini-3.5-flash", "name": "Google: Gemini 3.5 Flash", "short_name": "Gemini 3.5 Flash", "description": "Google's speed-focused Gemini Flash model for frontier multimodal intelligence across text, images, audio, video, PDFs, and code. Built for agentic coding, reliable tool use, structured outputs, and long-context workflows.", "context_length": 1000000, "created": 1779193800, "owned_by": "google", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 1.42 }, "output": { "unit": "1M tokens", "price_per_million_t": 8.55 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0.15 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "stop", "temperature", "top_p", "tool_choice", "tools" ], "capabilities": { "vision": true, "function_call": true, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": 55.3, "artificial_analysis_coding_index": 45, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": 0.922, "hle": 0.41, "livecodebench": null, "scicode": 0.531, "math_500": null, "aime": null, "aime_25": null, "ifbench": 0.763265306122449, "lcr": 0.693333333333333, "terminalbench_hard": 0.409090909090909, "tau2": 0.953216374269006 }, "discount": { "percentage": 5, "note": "A 5% discount applies to input and output token pricing only." } }, { "id": "gemma-4-31b-meromero", "name": "Google: Gemma 4 31B MeroMero", "short_name": "Gemma 4 31B MeroMero", "description": "Finetune for emotive dialogue, relationship scenes, and multimodal roleplay.", "context_length": 262144, "created": 1777593600, "owned_by": "google", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.28 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.28 } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "stop", "temperature", "top_p" ], "capabilities": { "vision": true, "function_call": false, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": null, "artificial_analysis_coding_index": null, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": null, "hle": null, "livecodebench": null, "scicode": null, "math_500": null, "aime": null, "aime_25": null, "ifbench": null, "lcr": null, "terminalbench_hard": null, "tau2": null }, "sub": true }, { "id": "gemma-4-31b-k1-v5", "name": "Google: Gemma 4 31B K1 v5", "short_name": "Gemma 4 31B K1 v5", "description": "Finetune for plot progression, action scenes, and multimodal roleplay.", "context_length": 262144, "created": 1777593600, "owned_by": "google", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.28 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.28 } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "stop", "temperature", "top_p" ], "capabilities": { "vision": true, "function_call": false, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": null, "artificial_analysis_coding_index": null, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": null, "hle": null, "livecodebench": null, "scicode": null, "math_500": null, "aime": null, "aime_25": null, "ifbench": null, "lcr": null, "terminalbench_hard": null, "tau2": null }, "sub": true }, { "id": "gemma-4-31b-larkspur-v0.5", "name": "Google: Gemma 4 31B Larkspur v0.5", "short_name": "Gemma 4 31B Larkspur v0.5", "description": "Finetune for character voice, slow-burn scenes, and multimodal roleplay", "context_length": 262144, "created": 1777593600, "owned_by": "google", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.28 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.28 } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "stop", "temperature", "top_p" ], "capabilities": { "vision": true, "function_call": false, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": null, "artificial_analysis_coding_index": null, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": null, "hle": null, "livecodebench": null, "scicode": null, "math_500": null, "aime": null, "aime_25": null, "ifbench": null, "lcr": null, "terminalbench_hard": null, "tau2": null }, "sub": true }, { "id": "gemma-4-31b-queen", "name": "Google: Gemma 4 31B Queen", "short_name": "Gemma 4 31B Queen", "description": "Finetune for commanding character voices, expressive dialogue, and multimodal chat", "context_length": 262144, "created": 1777593600, "owned_by": "google", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.28 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.28 } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "stop", "temperature", "top_p" ], "capabilities": { "vision": true, "function_call": false, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": null, "artificial_analysis_coding_index": null, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": null, "hle": null, "livecodebench": null, "scicode": null, "math_500": null, "aime": null, "aime_25": null, "ifbench": null, "lcr": null, "terminalbench_hard": null, "tau2": null }, "sub": true }, { "id": "gemma-4-31b-musica-v1", "name": "Google: Gemma 4 31B Musica v1", "short_name": "Gemma 4 31B Musica v1", "description": "Finetune for lyrical prose, theatrical scenes, and multimodal roleplay.", "context_length": 262144, "created": 1777593600, "owned_by": "google", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.28 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.28 } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "stop", "temperature", "top_p" ], "capabilities": { "vision": true, "function_call": false, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": null, "artificial_analysis_coding_index": null, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": null, "hle": null, "livecodebench": null, "scicode": null, "math_500": null, "aime": null, "aime_25": null, "ifbench": null, "lcr": null, "terminalbench_hard": null, "tau2": null }, "sub": true }, { "id": "gemma-4-31b-gemopus", "name": "Google: Gemma 4 31B Gemopus", "short_name": "Gemma 4 31B Gemopus", "description": "Finetune for reasoning-heavy story planning, branching scenes, and multimodal chat.", "context_length": 262144, "created": 1777593600, "owned_by": "google", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.28 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.28 } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "stop", "temperature", "top_p" ], "capabilities": { "vision": true, "function_call": false, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": null, "artificial_analysis_coding_index": null, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": null, "hle": null, "livecodebench": null, "scicode": null, "math_500": null, "aime": null, "aime_25": null, "ifbench": null, "lcr": null, "terminalbench_hard": null, "tau2": null }, "sub": true }, { "id": "gemma-4-31b-garnet-v2", "name": "Google: Gemma 4 31B Garnet V2", "short_name": "Gemma 4 31B Garnet V2", "description": "Finetune for polished prose, character consistency, and multimodal roleplay.", "context_length": 262144, "created": 1777593600, "owned_by": "google", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.28 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.28 } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "stop", "temperature", "top_p" ], "capabilities": { "vision": true, "function_call": false, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": null, "artificial_analysis_coding_index": null, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": null, "hle": null, "livecodebench": null, "scicode": null, "math_500": null, "aime": null, "aime_25": null, "ifbench": null, "lcr": null, "terminalbench_hard": null, "tau2": null }, "sub": true }, { "id": "gemma-4-31b-darkidol", "name": "Google: Gemma 4 31B DarkIdol", "short_name": "Gemma 4 31B DarkIdol", "description": "Finetune for dramatic tone, expressive dialogue, and multimodal roleplay.", "context_length": 262144, "created": 1777593600, "owned_by": "google", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.28 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.28 } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "stop", "temperature", "top_p" ], "capabilities": { "vision": true, "function_call": false, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": null, "artificial_analysis_coding_index": null, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": null, "hle": null, "livecodebench": null, "scicode": null, "math_500": null, "aime": null, "aime_25": null, "ifbench": null, "lcr": null, "terminalbench_hard": null, "tau2": null }, "sub": true }, { "id": "gemma-4-31b-meromero", "name": "Google: Gemma 4 31B MeroMero", "short_name": "Gemma 4 31B MeroMero", "description": "Finetune for emotive dialogue, relationship scenes, and multimodal roleplay.", "context_length": 262144, "created": 1777593600, "owned_by": "google", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.28 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.28 } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "stop", "temperature", "top_p" ], "capabilities": { "vision": true, "function_call": false, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": null, "artificial_analysis_coding_index": null, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": null, "hle": null, "livecodebench": null, "scicode": null, "math_500": null, "aime": null, "aime_25": null, "ifbench": null, "lcr": null, "terminalbench_hard": null, "tau2": null }, "sub": true }, { "id": "gemma-4-31b-claude-4.6-opus-reasoning-distilled", "name": "Google: Gemma 4 31B Claude 4.6 Opus Reasoning Distilled", "short_name": "Gemma 4 31B Claude 4.6 Opus Reasoning Distilled", "description": "Gemma‑4‑31B‑Claude‑4.6‑Opus‑Reasoning‑Distilled is an open‑source, 31‑billion‑parameter language model that has been fine‑tuned to “think” more like Anthropic’s top‑tier Claude 4.6 Opus, especially for complex reasoning, math, and coding tasks.", "context_length": 262144, "created": 1777593600, "owned_by": "google", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.28 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.28 } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "stop", "temperature", "top_p" ], "capabilities": { "vision": true, "function_call": false, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": null, "artificial_analysis_coding_index": null, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": null, "hle": null, "livecodebench": null, "scicode": null, "math_500": null, "aime": null, "aime_25": null, "ifbench": null, "lcr": null, "terminalbench_hard": null, "tau2": null }, "sub": true }, { "id": "gemma-4-31b-darkidol", "name": "Google: Gemma 4 31B Cognitive Unshackled", "short_name": "Gemma 4 31B Cognitive Unshackled", "description": "Finetune for open-ended reasoning, character chat, and multimodal creative work.", "context_length": 262144, "created": 1777593600, "owned_by": "google", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.28 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.28 } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "stop", "temperature", "top_p" ], "capabilities": { "vision": true, "function_call": false, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": null, "artificial_analysis_coding_index": null, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": null, "hle": null, "livecodebench": null, "scicode": null, "math_500": null, "aime": null, "aime_25": null, "ifbench": null, "lcr": null, "terminalbench_hard": null, "tau2": null }, "sub": true }, { "id": "ernie-5.1", "name": "Baidu: ERNIE 5.1", "short_name": "ERNIE 5.1", "description": "ERNIE 5.1 is Baidu's latest Wenxin model, with broad upgrades to agentic workflows, knowledge, reasoning, deep search, and multimodal understanding", "context_length": 119000, "created": 1778371200, "owned_by": "baidu", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.67 }, "output": { "unit": "1M tokens", "price_per_million_t": 2.7 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "stop", "temperature", "top_p", "tool_choice", "tools" ], "capabilities": { "vision": true, "function_call": true, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": null, "artificial_analysis_coding_index": null, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": null, "hle": null, "livecodebench": null, "scicode": null, "math_500": null, "aime": null, "aime_25": null, "ifbench": null, "lcr": null, "terminalbench_hard": null, "tau2": null } }, { "id": "gpt-5.5", "name": "OpenAI: GPT 5.5", "short_name": "GPT 5.5", "description": "GPT-5.5 is OpenAI's smartest and most intuitive model yet, built for agentic coding, computer use, and professional knowledge work with stronger reasoning and token efficiency.", "context_length": 1000000, "created": 1776902400, "owned_by": "openai", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 4.75 }, "output": { "unit": "1M tokens", "price_per_million_t": 28.5 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0.5 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "stop", "temperature", "top_p", "tool_choice", "tools" ], "capabilities": { "vision": true, "function_call": true, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": 60.2, "artificial_analysis_coding_index": 59.1, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": 0.935, "hle": 0.443, "livecodebench": null, "scicode": 0.561, "math_500": null, "aime": null, "aime_25": null, "ifbench": 0.759, "lcr": 0.743, "terminalbench_hard": 0.606, "tau2": 0.939 }, "discount": { "percentage": 5, "note": "A 5% discount applies to input and output token pricing only." } }, { "id": "granite-4.1-8b", "name": "IBM: Granite 4.1 8B", "short_name": "Granite 4.1 8B", "description": "IBM's Granite 4.1 8B is a dense, decoder-only 8-billion-parameter instruction model built for enterprise text workflows, including tool calling, retrieval-augmented generation, code generation with fill-in-the-middle support, summarization, classification, extraction, and multilingual assistance.", "context_length": 262144, "created": 1777420800, "owned_by": "ibm", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.04 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.09 } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "stop", "temperature", "top_p", "tool_choice", "tools" ], "capabilities": { "vision": true, "function_call": true, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": 12.4, "artificial_analysis_coding_index": 7.3, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": 0.433, "hle": 0.038, "livecodebench": null, "scicode": 0.218, "math_500": null, "aime": null, "aime_25": null, "ifbench": 0.386, "lcr": 0.12, "terminalbench_hard": 0, "tau2": 0.278 }, "sub": true }, { "id": "qwen3.5-27b-rprmax-v1", "name": "Qwen: Qwen 3.5 27B RpRMax v1", "short_name": "Qwen 3.5 27B RpRMax v1", "description": "RpRMax v1 finetune for roleplay-focused Qwen3.5 27B conversations and story generation.", "context_length": 262144, "created": 1777507200, "owned_by": "qwen", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.28 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.28 } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "stop", "temperature", "top_p" ], "capabilities": { "vision": true, "function_call": false, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": null, "artificial_analysis_coding_index": null, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": null, "hle": null, "livecodebench": null, "scicode": null, "math_500": null, "aime": null, "aime_25": null, "ifbench": null, "lcr": null, "terminalbench_hard": null, "tau2": null }, "sub": true }, { "id": "qwen3.5-27b-nanovel-derestricted", "name": "Qwen: Qwen 3.5 27B NaNovel Derestricted", "short_name": "Qwen 3.5 27B NaNovel Derestricted", "description": "NaNovel finetune for derestricted novel-style prose, character writing, and long-form scenes.", "context_length": 262144, "created": 1777507200, "owned_by": "qwen", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.28 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.28 } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "stop", "temperature", "top_p" ], "capabilities": { "vision": true, "function_call": false, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": null, "artificial_analysis_coding_index": null, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": null, "hle": null, "livecodebench": null, "scicode": null, "math_500": null, "aime": null, "aime_25": null, "ifbench": null, "lcr": null, "terminalbench_hard": null, "tau2": null }, "sub": true }, { "id": "qwen3.5-27b-marvin-dpo-v2-derestricted-lite", "name": "Qwen: Qwen 3.5 27B Marvin DPO V2 Derestricted Lite", "short_name": "Qwen 3.5 27B Marvin DPO V2 Derestricted Lite", "description": "Lighter Marvin DPO V2 finetune for responsive creative chat and iterative roleplay.", "context_length": 262144, "created": 1777507200, "owned_by": "qwen", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.28 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.28 } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "stop", "temperature", "top_p" ], "capabilities": { "vision": true, "function_call": false, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": null, "artificial_analysis_coding_index": null, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": null, "hle": null, "livecodebench": null, "scicode": null, "math_500": null, "aime": null, "aime_25": null, "ifbench": null, "lcr": null, "terminalbench_hard": null, "tau2": null }, "sub": true }, { "id": "qwen3.5-27b-marvin-dpo-v2-derestricted", "name": "Qwen: Qwen 3.5 27B Marvin DPO V2 Derestricted", "short_name": "Qwen 3.5 27B Marvin DPO V2 Derestricted", "description": "Marvin DPO V2 finetune for derestricted creative writing, dialogue, and roleplay.", "context_length": 262144, "created": 1777507200, "owned_by": "qwen", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.28 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.28 } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "stop", "temperature", "top_p" ], "capabilities": { "vision": true, "function_call": false, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": null, "artificial_analysis_coding_index": null, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": null, "hle": null, "livecodebench": null, "scicode": null, "math_500": null, "aime": null, "aime_25": null, "ifbench": null, "lcr": null, "terminalbench_hard": null, "tau2": null }, "sub": true }, { "id": "qwen3.5-27b-infracelestial", "name": "Qwen: Qwen 3.5 27B Infracelestial", "short_name": "Qwen 3.5 27B Infracelestial", "description": "Qwen 3.5 27B Infracelestial finetune for expressive creative chat and long-form roleplay.", "context_length": 262144, "created": 1777507200, "owned_by": "qwen", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.28 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.28 } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "stop", "temperature", "top_p" ], "capabilities": { "vision": true, "function_call": false, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": null, "artificial_analysis_coding_index": null, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": null, "hle": null, "livecodebench": null, "scicode": null, "math_500": null, "aime": null, "aime_25": null, "ifbench": null, "lcr": null, "terminalbench_hard": null, "tau2": null }, "sub": true }, { "id": "qwen3.5-27b-anko", "name": "Qwen: Qwen 3.5 27B Anko", "short_name": "Qwen 3.5 27B Anko", "description": "Qwen 3.5 27B Anko finetune for creative writing, roleplay, and multimodal chat.", "context_length": 262144, "created": 1777507200, "owned_by": "qwen", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.28 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.28 } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "stop", "temperature", "top_p" ], "capabilities": { "vision": true, "function_call": false, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": null, "artificial_analysis_coding_index": null, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": null, "hle": null, "livecodebench": null, "scicode": null, "math_500": null, "aime": null, "aime_25": null, "ifbench": null, "lcr": null, "terminalbench_hard": null, "tau2": null } }, { "id": "qwen3.5-27b-bluestar-v3-derestricted-lite", "name": "Qwen: Qwen 3.5 27B BlueStar v3 Derestricted Lite", "short_name": "Qwen 3.5 27B BlueStar v3 Derestricted Lite", "description": "Lighter third-generation BlueStar finetune for responsive creative chat, roleplay, and scene drafting.", "context_length": 262144, "created": 1777507200, "owned_by": "qwen", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.28 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.28 } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "stop", "temperature", "top_p" ], "capabilities": { "vision": true, "function_call": false, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": null, "artificial_analysis_coding_index": null, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": null, "hle": null, "livecodebench": null, "scicode": null, "math_500": null, "aime": null, "aime_25": null, "ifbench": null, "lcr": null, "terminalbench_hard": null, "tau2": null }, "sub": true }, { "id": "qwen3.5-27b-bluestar-v3-derestricted", "name": "Qwen: Qwen 3.5 27B BlueStar v3 Derestricted", "short_name": "Qwen 3.5 27B BlueStar v3 Derestricted", "description": "Third-generation BlueStar finetune for creative roleplay, narrative prose, and multimodal Qwen3.5 27B workflows.", "context_length": 262144, "created": 1777507200, "owned_by": "qwen", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.28 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.28 } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "stop", "temperature", "top_p" ], "capabilities": { "vision": true, "function_call": false, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": null, "artificial_analysis_coding_index": null, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": null, "hle": null, "livecodebench": null, "scicode": null, "math_500": null, "aime": null, "aime_25": null, "ifbench": null, "lcr": null, "terminalbench_hard": null, "tau2": null }, "sub": true }, { "id": "grok-4.3", "name": "xAI: Grok 4.3", "short_name": "Grok 4.3", "description": "Grok 4.3 is xAI's reasoning model for text and image inputs, built for agentic workflows, instruction following, factual accuracy, long-document analysis, and deep research.", "context_length": 1000000, "created": 1777507200, "owned_by": "xai", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 1.19 }, "output": { "unit": "1M tokens", "price_per_million_t": 2.38 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0.2 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "stop", "temperature", "top_p", "tools", "tool_choice" ], "capabilities": { "vision": true, "function_call": true, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": 53.2, "artificial_analysis_coding_index": 41, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": 0.901, "hle": 0.35, "livecodebench": null, "scicode": 0.473, "math_500": null, "aime": null, "aime_25": null, "ifbench": 0.813, "lcr": 0.643, "terminalbench_hard": 0.379, "tau2": 0.977 } }, { "id": "deepseek-v4-pro", "name": "DeepSeek: DeepSeek V4 Pro", "short_name": "DeepSeek V4 Pro", "description": "DeepSeek V4 Pro is a large-scale Mixture-of-Experts model from DeepSeek with a 1M-token context window, built for advanced reasoning, coding, long-horizon agent workflows, knowledge, math, and software engineering tasks.", "context_length": 1048576, "created": 1776988800, "owned_by": "deepseek-ai", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 1.57 }, "output": { "unit": "1M tokens", "price_per_million_t": 3.13 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0.22 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "stop", "temperature", "top_p", "tools", "tool_choice" ], "capabilities": { "vision": false, "function_call": true, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": null, "artificial_analysis_coding_index": null, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": 0.888, "hle": 0.359, "livecodebench": null, "scicode": 0.5, "math_500": null, "aime": null, "aime_25": null, "ifbench": 0.765, "lcr": 0.663, "terminalbench_hard": null, "tau2": 0.962 }, "sub": true }, { "id": "deepseek-v4-flash", "name": "DeepSeek: DeepSeek V4 Flash", "short_name": "DeepSeek V4 Flash", "description": "DeepSeek V4 Flash is an efficiency-optimized Mixture-of-Experts model from DeepSeek with a 1M-token context window, built for fast inference, high-throughput workloads, reasoning, coding, and agent workflows.", "context_length": 1048576, "created": 1776988800, "owned_by": "deepseek-ai", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.12 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.25 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0.045 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "stop", "temperature", "top_p", "tools", "tool_choice" ], "capabilities": { "vision": false, "function_call": true, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": 44.9, "artificial_analysis_coding_index": 39.8, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": 0.867, "hle": 0.278, "livecodebench": null, "scicode": 0.42, "math_500": null, "aime": null, "aime_25": null, "ifbench": 0.735, "lcr": 0.627, "terminalbench_hard": 0.386, "tau2": 0.956 }, "sub": true }, { "id": "qwen3.6-27b", "name": "Qwen: Qwen3.6 27B", "short_name": "Qwen3.6 27B", "description": "Qwen3.6‑27B is a dense multimodal large language model from Qwen team, optimized for agentic coding and general reasoning tasks. It excels on code‑agent benchmarks and supports long‑context interactions, making it suitable for complex programming and reasoning workflows.", "context_length": 260096, "created": 1776902400, "owned_by": "qwen", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.54 }, "output": { "unit": "1M tokens", "price_per_million_t": 3.24 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0.3 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "stop", "temperature", "top_p" ], "capabilities": { "vision": true, "function_call": true, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": null, "artificial_analysis_coding_index": null, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": null, "hle": null, "livecodebench": null, "scicode": null, "math_500": null, "aime": null, "aime_25": null, "ifbench": null, "lcr": null, "terminalbench_hard": null, "tau2": null } }, { "id": "mimo-v2.5", "name": "Xiaomi: MiMo V2.5", "short_name": "MiMo V2.5", "description": "MiMo V2.5 is Xiaomi's full-modal understanding model for agent workflows. It supports deep reasoning, tool calling, structured outputs, and web search with up to 1M context.", "context_length": 1048576, "created": 1776816000, "owned_by": "xiaomi", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.14 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.28 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0.0028 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "stop", "temperature", "top_p" ], "capabilities": { "vision": true, "function_call": true, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": null, "artificial_analysis_coding_index": null, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": null, "hle": null, "livecodebench": null, "scicode": null, "math_500": null, "aime": null, "aime_25": null, "ifbench": null, "lcr": null, "terminalbench_hard": null, "tau2": null }, "sub": true }, { "id": "mimo-v2.5-pro", "name": "Xiaomi: MiMo V2.5 Pro", "short_name": "MiMo V2.5 Pro", "description": "MiMo V2.5 Pro is Xiaomi's long-context flagship general model for coding and agentic orchestration. It supports reasoning, tool calling, and structured outputs with up to 1M context.", "context_length": 1048576, "created": 1776816000, "owned_by": "xiaomi", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.435 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.87 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0.0036 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "stop", "temperature", "top_p" ], "capabilities": { "vision": true, "function_call": true, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": 53.8, "artificial_analysis_coding_index": 45.5, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": 0.866, "hle": 0.338, "livecodebench": null, "scicode": 0.502, "math_500": null, "aime": null, "aime_25": null, "ifbench": 0.799, "lcr": 0.733, "terminalbench_hard": 0.432, "tau2": 0.942 }, "sub": true }, { "id": "qwen3.5-27b-bluestar-v2-derestricted", "name": "Qwen: Qwen3.5-27B BlueStar v2 Derestricted", "short_name": "Qwen3.5-27B BlueStar v2 Derestricted", "description": "Second-generation BlueStar finetune for more polished prose, character voice, and open-ended roleplay.", "context_length": 262144, "created": 1775433600, "owned_by": "qwen", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.28 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.28 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "stop", "temperature", "top_p" ], "capabilities": { "vision": true, "function_call": false, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": null, "artificial_analysis_coding_index": null, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": null, "hle": null, "livecodebench": null, "scicode": null, "math_500": null, "aime": null, "aime_25": null, "ifbench": null, "lcr": null, "terminalbench_hard": null, "tau2": null }, "sub": true }, { "id": "qwen3.5-27b-bluestar-derestricted-lite", "name": "Qwen: Qwen3.5-27B BlueStar v2 Derestricted", "short_name": "Qwen3.5-27B BlueStar v2 Derestricted", "description": "Lighter-tuned BlueStar variant aimed at responsive creative roleplay and storytelling while keeping the same Qwen3.5 27B multimodal base.", "context_length": 262144, "created": 1775433600, "owned_by": "qwen", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.28 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.28 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "stop", "temperature", "top_p" ], "capabilities": { "vision": true, "function_call": false, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": null, "artificial_analysis_coding_index": null, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": null, "hle": null, "livecodebench": null, "scicode": null, "math_500": null, "aime": null, "aime_25": null, "ifbench": null, "lcr": null, "terminalbench_hard": null, "tau2": null }, "sub": true }, { "id": "qwen3.5-27b-bluestar-derestricted", "name": "Qwen: Qwen3.5-27B BlueStar v2 Derestricted", "short_name": "Qwen3.5-27B BlueStar v2 Derestricted", "description": "Creative Qwen3.5 27B finetune tuned for expressive roleplay and high-energy storytelling.", "context_length": 262144, "created": 1775433600, "owned_by": "qwen", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.28 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.28 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "stop", "temperature", "top_p" ], "capabilities": { "vision": true, "function_call": false, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": null, "artificial_analysis_coding_index": null, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": null, "hle": null, "livecodebench": null, "scicode": null, "math_500": null, "aime": null, "aime_25": null, "ifbench": null, "lcr": null, "terminalbench_hard": null, "tau2": null }, "sub": true }, { "id": "qwen3.5-27b-vivid-durian", "name": "Qwen: Qwen3.5 27B Vivid Durian", "short_name": "Qwen3.5 27B Vivid Durian", "description": "Creative Qwen3.5 27B finetune optimized for vivid, expressive writing and roleplay.", "context_length": 262144, "created": 1773792000, "owned_by": "qwen", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.28 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.28 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "stop", "temperature", "top_p" ], "capabilities": { "vision": true, "function_call": false, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": null, "artificial_analysis_coding_index": null, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": null, "hle": null, "livecodebench": null, "scicode": null, "math_500": null, "aime": null, "aime_25": null, "ifbench": null, "lcr": null, "terminalbench_hard": null, "tau2": null }, "sub": true }, { "id": "qwen3.5-27b-musica-v1", "name": "Qwen: Qwen3.5 27B Musica v1", "short_name": "Qwen3.5 27B Musica v1", "description": "Creative Qwen3.5 27B roleplay, story generation, and conversational finetune built on ArliAI's derestricted base with reasoning and vision support.", "context_length": 262144, "created": 1774569600, "owned_by": "qwen", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.28 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.28 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "stop", "temperature", "top_p" ], "capabilities": { "vision": true, "function_call": false, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": null, "artificial_analysis_coding_index": null, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": null, "hle": null, "livecodebench": null, "scicode": null, "math_500": null, "aime": null, "aime_25": null, "ifbench": null, "lcr": null, "terminalbench_hard": null, "tau2": null }, "sub": true }, { "id": "ling-2.6-flash:free", "name": "inclusionAI: Ling 2.6 Flash (Free)", "short_name": "Ling 2.6 Flash (Free)", "description": "Ling 2.6 Flash is an instruct model from inclusionAI with 104B total parameters and 7.4B active parameters, designed for real-world agents that require fast responses, strong execution, and high token efficiency. It delivers performance comparable to state-of-the-art models at a similar scale while significantly reducing token usage across coding, document processing, and lightweight agent workflows.", "context_length": 262144, "created": 1776795886, "owned_by": "inclusionai", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0 }, "output": { "unit": "1M tokens", "price_per_million_t": 0 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": false, "function_call": true, "reasoning": false }, "benchmarks": { "artificial_analysis_intelligence_index": null, "artificial_analysis_coding_index": null, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": null, "hle": null, "livecodebench": null, "scicode": null, "math_500": null, "aime": null, "aime_25": null, "ifbench": null, "lcr": null, "terminalbench_hard": null, "tau2": null } }, { "id": "qwen3.6-flash", "name": "Qwen: Qwen3.6 Flash", "short_name": "Qwen3.6 Flash", "description": "Qwen3.6 Flash is Alibaba's fast native vision-language model in the Qwen 3.6 family. It improves over 3.5 Flash with stronger coding/agent performance and better spatial intelligence, including object localization and detection.", "context_length": 991808, "created": 1776384000, "owned_by": "qwen", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.18 }, "output": { "unit": "1M tokens", "price_per_million_t": 1.04 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0.02 }, "write": { "5m": { "unit": "1M tokens", "price_per_million_t": 0.24 }, "1h": { "unit": "1M tokens", "price_per_million_t": 0.38 } } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": true, "function_call": true, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": null, "artificial_analysis_coding_index": null, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": null, "hle": null, "livecodebench": null, "scicode": null, "math_500": null, "aime": null, "aime_25": null, "ifbench": null, "lcr": null, "terminalbench_hard": null, "tau2": null }, "sub": true }, { "id": "qwen3.6-35b-a3b", "name": "Qwen: Qwen3.6 35B A3B", "short_name": "Qwen3.6 35B A3B", "description": "Qwen3.6 35B A3B is a native vision-language MoE model with hybrid attention. Compared to Qwen3.5 35B A3B, Alibaba reports stronger agentic coding, mathematical and code reasoning, and better spatial understanding (including object localization and detection).", "context_length": 262144, "created": 1776384000, "owned_by": "qwen", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.27 }, "output": { "unit": "1M tokens", "price_per_million_t": 1.57 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0.12 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": true, "function_call": true, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": 43.5, "artificial_analysis_coding_index": 35.1, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": 0.841, "hle": 0.202, "livecodebench": null, "scicode": 0.358, "math_500": null, "aime": null, "aime_25": null, "ifbench": 0.644, "lcr": 0.637, "terminalbench_hard": 0.348, "tau2": 0.953 } }, { "id": "qwen3.6-max-preview", "name": "Qwen: Qwen3.6 Max Preview", "short_name": "Qwen3.6 Max Preview", "description": "Qwen3.6-Max-Preview is Alibaba's flagship preview model with over 1 trillion parameters. It supports 262K token context, 100+ languages, and excels in agentic coding, multi-file analysis, and long-context tasks", "context_length": 245760, "created": 1776643200, "owned_by": "qwen", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 1.17 }, "output": { "unit": "1M tokens", "price_per_million_t": 7.41 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0.12 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": false, "function_call": true, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": 51.8, "artificial_analysis_coding_index": 44.9, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": 0.888, "hle": 0.289, "livecodebench": null, "scicode": 0.469, "math_500": null, "aime": null, "aime_25": null, "ifbench": 0.766, "lcr": 0.697, "terminalbench_hard": 0.439, "tau2": 0.959 } }, { "id": "gemma-4-31b-it:free", "name": "Google: Gemma 4 31B (Free)", "short_name": "Gemma 4 31B (Free)", "description": "Google's Gemma 4 31B instruction-tuned model for heavier reasoning, coding, agentic workflows, and long-context multimodal understanding. ", "context_length": 262144, "created": 1775088000, "owned_by": "google", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0 }, "output": { "unit": "1M tokens", "price_per_million_t": 0 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": true, "function_call": true, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": 39.2, "artificial_analysis_coding_index": 38.7, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": 0.857, "hle": 0.227, "livecodebench": null, "scicode": 0.434, "math_500": null, "aime": null, "aime_25": null, "ifbench": 0.756, "lcr": 0.62, "terminalbench_hard": 0.364, "tau2": 0.599 }, "note": "available only on paid plan" }, { "id": "gemma-4-26b-a4b-it", "name": "Google: Gemma 4 26B A4B", "short_name": "Gemma 4 26B A4B", "description": "Google's Gemma 4 26B A4B instruction-tuned model built for scalable reasoning, coding, long-context, and multimodal workflows.", "context_length": 262144, "created": 1775088000, "owned_by": "google", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.11 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.36 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0.06 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": true, "function_call": true, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": 31.2, "artificial_analysis_coding_index": 22.4, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": 0.792, "hle": 0.183, "livecodebench": null, "scicode": 0.4, "math_500": null, "aime": null, "aime_25": null, "ifbench": 0.724, "lcr": 0.557, "terminalbench_hard": 0.136, "tau2": 0.436 } }, { "id": "gemma-4-31b-it", "name": "Google: Gemma 4 31B", "short_name": "Gemma 4 31B", "description": "Google's Gemma 4 31B instruction-tuned model for heavier reasoning, coding, agentic workflows, and long-context multimodal understanding. ", "context_length": 262144, "created": 1775088000, "owned_by": "google", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.11 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.33 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0.09 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": true, "function_call": true, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": 39.2, "artificial_analysis_coding_index": 38.7, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": 0.857, "hle": 0.227, "livecodebench": null, "scicode": 0.434, "math_500": null, "aime": null, "aime_25": null, "ifbench": 0.756, "lcr": 0.62, "terminalbench_hard": 0.364, "tau2": 0.599 }, "sub": true }, { "id": "kimi-k2.6", "name": "MoonshotAI: Kimi K2.6", "short_name": "Kimi K2.6", "description": "Kimi K2.6 Code is currently in preview/testing and is built for long-context coding, tool use, and agent-style workflows.", "context_length": 1000000, "created": 1776297600, "owned_by": "moonshotai", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.57 }, "output": { "unit": "1M tokens", "price_per_million_t": 2.85 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0.19 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": true, "function_call": true, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": 53.9, "artificial_analysis_coding_index": 47.1, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": 0.911, "hle": 0.359, "livecodebench": null, "scicode": 0.535, "math_500": null, "aime": null, "aime_25": null, "ifbench": 0.76, "lcr": 0.697, "terminalbench_hard": 0.439, "tau2": 0.959 } }, { "id": "claude-opus-4.7", "name": "Anthropic: Claude Opus 4.7", "short_name": "Claude Opus 4.7", "description": "Claude Opus 4.7 is a major upgrade for advanced software engineering, long-running complex tasks, and high-resolution vision understanding.", "context_length": 1000000, "created": 1776297600, "owned_by": "anthropic", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 4.75 }, "output": { "unit": "1M tokens", "price_per_million_t": 23.75 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0.5 }, "write": { "5m": { "unit": "1M tokens", "price_per_million_t": 6.25 }, "1h": { "unit": "1M tokens", "price_per_million_t": 10 } } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": true, "function_call": true, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": 57.3, "artificial_analysis_coding_index": 52.5, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": 0.914, "hle": 0.396, "livecodebench": null, "scicode": 0.545, "math_500": null, "aime": null, "aime_25": null, "ifbench": 0.586, "lcr": 0.703, "terminalbench_hard": 0.515, "tau2": 0.886 }, "discount": { "percentage": 5, "note": "A 5% discount applies to input and output token pricing only." } }, { "id": "glm-5.1", "name": "Z.ai: GLM 5.1", "short_name": "GLM 5.1", "description": "GLM-5.1 is Zhipu's next-level open source model, ranking #1 in open source and #3 globally across SWE-Bench Pro, Terminal-Bench, and NL2Repo (as of April 2026). Built for long-horizon tasks, it can run autonomously for up to 8 hours.", "context_length": 202752, "created": 1775578025, "owned_by": "z-ai", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.83 }, "output": { "unit": "1M tokens", "price_per_million_t": 2.48 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0.28 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": false, "function_call": true, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": 51.4, "artificial_analysis_coding_index": 43.4, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": 0.868, "hle": 0.28, "livecodebench": null, "scicode": 0.438, "math_500": null, "aime": null, "aime_25": null, "ifbench": 0.763, "lcr": 0.623, "terminalbench_hard": 0.432, "tau2": 0.977 } }, { "id": "step-3.5-flash", "name": "StepFun: Step 3.5 Flash", "short_name": "Step 3.5 Flash", "description": "StepFun's most capable open-source reasoning model with visible reasoning traces. Built on a sparse Mixture-of-Experts architecture with 196B total parameters and only 11B active per token, it achieves frontier-level performance in math, logic, and agentic coding", "context_length": 256000, "created": 1769990400, "owned_by": "stepfun", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.1 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.3 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0.02 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": false, "function_call": false, "reasoning": false }, "benchmarks": { "artificial_analysis_intelligence_index": 37.8, "artificial_analysis_coding_index": 31.6, "artificial_analysis_math_index": null, "mmlu_pro": 0.338, "gpqa": 0.831, "hle": 0.191, "livecodebench": 0.048, "scicode": 0.404, "math_500": 0.164, "aime": 0.007, "aime_25": null, "ifbench": 0.646, "lcr": 0.43, "terminalbench_hard": 0.273, "tau2": 0.944 } }, { "id": "step-3.5-flash:free", "name": "StepFun: Step 3.5 Flash (Free)", "short_name": "Step 3.5 Flash (Free)", "description": "StepFun's most capable open-source reasoning model with visible reasoning traces. Built on a sparse Mixture-of-Experts architecture with 196B total parameters and only 11B active per token, it achieves frontier-level performance in math, logic, and agentic coding", "context_length": 256000, "created": 1769990400, "owned_by": "stepfun", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0 }, "output": { "unit": "1M tokens", "price_per_million_t": 0 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": false, "function_call": false, "reasoning": false }, "benchmarks": { "artificial_analysis_intelligence_index": 37.8, "artificial_analysis_coding_index": 31.6, "artificial_analysis_math_index": null, "mmlu_pro": 0.338, "gpqa": 0.831, "hle": 0.191, "livecodebench": 0.048, "scicode": 0.404, "math_500": 0.164, "aime": 0.007, "aime_25": null, "ifbench": 0.646, "lcr": 0.43, "terminalbench_hard": 0.273, "tau2": 0.944 } }, { "id": "cydonia-24B-v4.3", "name": "The Drummer: Cydonia 24B v4.3", "short_name": "Cydonia 24B v4.3", "description": "The model is specifically optimized for uncensored roleplay, character portrayal, and collaborative storytelling. It stands out for its ability to take initiative in narratives — organically introducing story elements, maintaining distinct character personalities across long contexts, and handling multi-character group chats with consistency.", "context_length": 32768, "created": 1766620800, "owned_by": "thedrummer", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.09 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.12 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": false, "function_call": false, "reasoning": false }, "benchmarks": { "artificial_analysis_intelligence_index": null, "artificial_analysis_coding_index": null, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": null, "hle": null, "livecodebench": null, "scicode": null, "math_500": null, "aime": null, "aime_25": null, "ifbench": null, "lcr": null, "terminalbench_hard": null, "tau2": null } }, { "id": "gpt-5.4", "name": "OpenAI: GPT 5.4", "short_name": "GPT 5.4", "description": "GPT-5.4 is OpenAI's latest frontier model for professional work with stronger reasoning, coding, and tool use.", "context_length": 400000, "created": 1772668800, "owned_by": "openai", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 2.38 }, "output": { "unit": "1M tokens", "price_per_million_t": 14.25 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0.25 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": true, "function_call": true, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": 57.2, "artificial_analysis_coding_index": 57.3, "artificial_analysis_math_index": null, "mmlu_pro": 0.338, "gpqa": 0.92, "hle": 0.416, "livecodebench": 0.048, "scicode": 0.566, "math_500": 0.164, "aime": 0.007, "aime_25": null, "ifbench": 0.739, "lcr": 0.74, "terminalbench_hard": 0.576, "tau2": 0.915 }, "discount": { "percentage": 5, "note": "A 5% discount applies to input and output token pricing only." } }, { "id": "gpt-5.4-mini", "name": "OpenAI: GPT 5.4 Mini", "short_name": "GPT 5.4 Mini", "description": "GPT-5.4 mini brings the core capabilities of GPT-5.4 to a faster, more efficient model optimized for high-throughput workloads.", "context_length": 400000, "created": 1773705600, "owned_by": "openai", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.71 }, "output": { "unit": "1M tokens", "price_per_million_t": 4.28 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0.08 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": true, "function_call": true, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": 48.1, "artificial_analysis_coding_index": 51.5, "artificial_analysis_math_index": null, "mmlu_pro": 0.338, "gpqa": 0.875, "hle": 0.266, "livecodebench": 0.048, "scicode": 0.499, "math_500": 0.164, "aime": 0.007, "aime_25": null, "ifbench": 0.733, "lcr": 0.693, "terminalbench_hard": 0.523, "tau2": 0.737 }, "discount": { "percentage": 5, "note": "A 5% discount applies to input and output token pricing only." } }, { "id": "gpt-5.4-nano", "name": "OpenAI: GPT 5.4 Nano", "short_name": "GPT 5.4 Nano", "description": "GPT-5.4 nano is the most lightweight and cost-efficient variant of the GPT-5.4 family, optimized for speed-critical and high-volume tasks.", "context_length": 400000, "created": 1773705600, "owned_by": "openai", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.19 }, "output": { "unit": "1M tokens", "price_per_million_t": 1.19 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0.02 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": true, "function_call": true, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": 24.4, "artificial_analysis_coding_index": 27.9, "artificial_analysis_math_index": null, "mmlu_pro": 0.338, "gpqa": 0.558, "hle": 0.042, "livecodebench": 0.048, "scicode": 0.352, "math_500": 0.164, "aime": 0.007, "aime_25": null, "ifbench": 0.327, "lcr": 0.247, "terminalbench_hard": 0.242, "tau2": 0.348 }, "discount": { "percentage": 5, "note": "A 5% discount applies to input and output token pricing only." } }, { "id": "mercury-2", "name": "Inception Labs: Mercury 2", "short_name": "Mercury 2", "description": "Mercury 2 — the world's fastest reasoning language model, built to make production AI feel instant.", "context_length": 128000, "created": 1704067200, "owned_by": "inceptionlabs", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.23 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.69 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0.026 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": false, "function_call": true, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": 32.8, "artificial_analysis_coding_index": 30.6, "artificial_analysis_math_index": null, "mmlu_pro": 0.338, "gpqa": 0.77, "hle": 0.155, "livecodebench": 0.048, "scicode": 0.387, "math_500": 0.164, "aime": 0.007, "aime_25": null, "ifbench": 0.698, "lcr": 0.363, "terminalbench_hard": 0.265, "tau2": 0.708 } }, { "id": "qwen-image-2.0", "name": "Qwen: Qwen Image 2.0", "short_name": "Qwen Image 2.0", "description": "Qwen-Image-2.0 is Alibaba's next-generation AI image generation model that turns text descriptions into stunning, photorealistic visuals at native 2K resolution. It excels at rendering clean, readable text directly inside images — making it easy to generate professional-looking posters, infographics, slides, and comics without the garbled letters common in other AI image tools. Built on a compact 7B architecture, it's faster and more efficient than its predecessor while still delivering sharp, detailed results.", "created": 1738368000, "owned_by": "qwen", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/images/generations", "/v1/images/edits" ], "pricing": { "unit": "per image", "resolutions": { "1024x1024": 0.0297, "1280x720": 0.0297, "720x1280": 0.0297, "1536x1024": 0.0297, "1024x1536": 0.0297, "auto": 0.0297 } }, "supported_sizes": [ "1024x1024", "1280x720", "720x1280", "1536x1024", "1024x1536" ], "benchmarks": null }, { "id": "qwen-image-2.0-pro", "name": "Qwen: Qwen Image 2.0 Pro", "short_name": "Qwen Image 2.0 Pro", "description": "Qwen-Image-2.0-Pro is the premium version of Qwen-Image-2.0, built for creators and professionals who need the highest quality output. It delivers superior visual fidelity with richer detail — think lifelike hair textures, skin tones, fabric, and intricate typography — while also handling complex, multi-element prompts more accurately than the standard model. It's the go-to choice for production-ready work like advertising, branding, and high-impact visual content.", "created": 1738368000, "owned_by": "qwen", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/images/generations", "/v1/images/edits" ], "pricing": { "unit": "per image", "resolutions": { "1024x1024": 0.0693, "1280x720": 0.0693, "720x1280": 0.0693, "1536x1024": 0.0693, "1024x1536": 0.0693, "auto": 0.0693 } }, "supported_sizes": [ "1024x1024", "1280x720", "720x1280", "1536x1024", "1024x1536" ], "benchmarks": null }, { "id": "seedream-v5.0-lite", "name": "ByteDance: Seedream 5.0 Lite", "short_name": "Seedream 5.0 Lite", "description": "Seedream 5.0 Lite is ByteDance's latest AI image generation model that goes beyond just following your prompt — it actually reasons through what you're trying to create. It's the first image model with built-in real-time web search, so it can pull up-to-date information before generating a visual, making it great for trending topics or data-driven imagery like charts and infographics.", "created": 1738368000, "owned_by": "bytedance", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/images/generations", "/v1/images/edits" ], "pricing": { "unit": "per image", "resolutions": { "1440x2560": 0.0385, "2560x1440": 0.0385, "2048x2048": 0.0385, "3072x2048": 0.0385, "2048x3072": 0.0385, "4096x2304": 0.0385, "2304x4096": 0.0385, "auto": 0.0385 } }, "supported_sizes": [ "2048x2048", "2560x1440", "1440x2560", "3072x2048", "2048x3072", "4096x2304", "2304x4096" ], "benchmarks": { "elo": 1122, "rank": 30, "ci95": "-11/11", "appearances": 3508, "categories": { "style_general_photorealistic": { "elo": 1107, "ci95": "-21/21", "appearances": 1124 }, "style_cartoon_illustration": { "elo": 1198, "ci95": "-42/42", "appearances": 294 }, "subject_people:_portraits": { "elo": 1159, "ci95": "-38/38", "appearances": 357 }, "style_anime": { "elo": 1139, "ci95": "-45/45", "appearances": 243 }, "subject_futuristic_sci-fi": { "elo": 1072, "ci95": "-34/34", "appearances": 418 }, "subject_physical_spaces": { "elo": 1085, "ci95": "-34/34", "appearances": 411 }, "subject_text_typography": { "elo": 1105, "ci95": "-54/54", "appearances": 169 }, "subject_nature_landscapes": { "elo": 1100, "ci95": "-42/42", "appearances": 275 }, "style_vintage_retro": { "elo": 1106, "ci95": "-56/56", "appearances": 158 }, "subject_fantasy_mythical": { "elo": 1175, "ci95": "-32/32", "appearances": 466 }, "subject_commercial": { "elo": 1136, "ci95": "-45/45", "appearances": 243 }, "style_traditional_art": { "elo": 1182, "ci95": "-48/48", "appearances": 207 }, "style_graphic_design_digital_rendering": { "elo": 1093, "ci95": "-22/22", "appearances": 996 }, "subject_people:_groups_activities": { "elo": 1152, "ci95": "-39/39", "appearances": 324 } } } }, { "id": "gpt-5.1", "name": "OpenAI: GPT 5.1", "short_name": "GPT 5.1", "description": "GPT-5.1 is the latest frontier-grade model in the GPT-5 series, offering stronger general-purpose reasoning, improved instruction adherence, and a more natural conversational style compared to GPT-5. It uses adaptive reasoning to allocate computation dynamically, responding quickly to simple queries while spending more depth on complex tasks. The model produces clearer, more grounded explanations with reduced jargon, making it easier to follow even on technical or multi-step problems. Built for broad task coverage, GPT-5.1 delivers consistent gains across math, coding, and structured analysis workloads, with more coherent long-form answers and improved tool-use reliability. It also features refined conversational alignment, enabling warmer, more intuitive responses without compromising precision. GPT-5.1 serves as the primary full-capability successor to GPT-5", "context_length": 400000, "created": 1763060305, "owned_by": "openai", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 1.19 }, "output": { "unit": "1M tokens", "price_per_million_t": 9.5 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0.13 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": true, "function_call": true, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": 47.7, "artificial_analysis_coding_index": 44.7, "artificial_analysis_math_index": 94, "mmlu_pro": 0.87, "gpqa": 0.873, "hle": 0.265, "livecodebench": 0.868, "scicode": 0.433, "math_500": null, "aime": null, "aime_25": 0.94, "ifbench": 0.729, "lcr": 0.75, "terminalbench_hard": 0.455, "tau2": 0.819 }, "discount": { "percentage": 5, "note": "A 5% discount applies to input and output token pricing only." } }, { "id": "gpt-5.2", "name": "OpenAI: GPT 5.2", "short_name": "GPT 5.2", "description": "GPT-5.2 is the latest frontier-grade model in the GPT-5 series, offering stronger agentic and long context perfomance compared to GPT-5.1. It uses adaptive reasoning to allocate computation dynamically, responding quickly to simple queries while spending more depth on complex tasks. Built for broad task coverage, GPT-5.2 delivers consistent gains across math, coding, sciende, and tool calling workloads, with more coherent long-form answers and improved tool-use reliability.", "context_length": 400000, "created": 1765389775, "owned_by": "openai", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 1.66 }, "output": { "unit": "1M tokens", "price_per_million_t": 13.3 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0.18 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": true, "function_call": true, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": 51.3, "artificial_analysis_coding_index": 48.7, "artificial_analysis_math_index": 99, "mmlu_pro": 0.874, "gpqa": 0.903, "hle": 0.354, "livecodebench": 0.889, "scicode": 0.521, "math_500": null, "aime": null, "aime_25": 0.99, "ifbench": 0.754, "lcr": 0.727, "terminalbench_hard": 0.47, "tau2": 0.848 }, "discount": { "percentage": 5, "note": "A 5% discount applies to input and output token pricing only." } }, { "id": "gpt-5.3-codex", "name": "OpenAI: GPT 5.3 Codex", "short_name": "GPT 5.3 Codex", "description": "Coding-focused GPT-5.3 variant with optimized routing.", "context_length": 400000, "created": 1771891200, "owned_by": "openai", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 1.66 }, "output": { "unit": "1M tokens", "price_per_million_t": 13.3 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0.18 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": true, "function_call": true, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": 54, "artificial_analysis_coding_index": 53.1, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": 0.915, "hle": 0.399, "livecodebench": null, "scicode": 0.532, "math_500": null, "aime": null, "aime_25": null, "ifbench": 0.754, "lcr": 0.74, "terminalbench_hard": 0.53, "tau2": 0.909 }, "discount": { "percentage": 5, "note": "A 5% discount applies to input and output token pricing only." } }, { "id": "qwen3.5-122b-a10b", "name": "Qwen: Qwen 3.5 122B A10B", "short_name": "Qwen 3.5 122B A10B", "description": "Qwen 3.5 122B A10B is Alibaba's high-end native vision-language model in the Qwen 3.5 family with strong text and multimodal performance.", "context_length": 260096, "created": 1771891200, "owned_by": "qwen", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.37 }, "output": { "unit": "1M tokens", "price_per_million_t": 2.94 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": true, "function_call": false, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": 41.6, "artificial_analysis_coding_index": 34.7, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": 0.857, "hle": 0.234, "livecodebench": null, "scicode": 0.42, "math_500": null, "aime": null, "aime_25": null, "ifbench": 0.757, "lcr": 0.667, "terminalbench_hard": 0.311, "tau2": 0.936 } }, { "id": "qwen3.5-27b", "name": "Qwen: Qwen 3.5 27", "short_name": "Qwen 3.5 27", "description": "Qwen 3.5 27B is a native vision-language dense model optimized for fast responses while balancing quality and inference speed.", "context_length": 260096, "created": 1771891200, "owned_by": "qwen", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.28 }, "output": { "unit": "1M tokens", "price_per_million_t": 2.2 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": true, "function_call": false, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": 37.1, "artificial_analysis_coding_index": 30.3, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": 0.845, "hle": 0.197, "livecodebench": null, "scicode": 0.377, "math_500": null, "aime": null, "aime_25": null, "ifbench": 0.725, "lcr": 0.627, "terminalbench_hard": 0.265, "tau2": 0.892 }, "sub": true }, { "id": "qwen3.5-35b-a3b", "name": "Qwen: Qwen 3.5 35B A3B", "short_name": "Qwen 3.5 35B A3B", "description": "Qwen 3.5 35B A3B is a native vision-language MoE model with hybrid attention designed for efficient inference and strong general performance", "context_length": 260096, "created": 1771891200, "owned_by": "qwen", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.18 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.46 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0.11 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": true, "function_call": false, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": 37.1, "artificial_analysis_coding_index": 30.3, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": 0.845, "hle": 0.197, "livecodebench": null, "scicode": 0.377, "math_500": null, "aime": null, "aime_25": null, "ifbench": 0.725, "lcr": 0.627, "terminalbench_hard": 0.265, "tau2": 0.892 } }, { "id": "qwen3.5-flash", "name": "Qwen: Qwen 3.5 Flash", "short_name": "Qwen 3.5 Flash", "description": "Qwen 3.5 Flash is the fastest and cheapest native Qwen 3.5 vision-language model with a 1M-token context window.", "context_length": 991808, "created": 1771891200, "owned_by": "qwen", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.09 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.37 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0.02 }, "write": { "5m": { "unit": "1M tokens", "price_per_million_t": 0.24 }, "1h": { "unit": "1M tokens", "price_per_million_t": 0.38 } } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": true, "function_call": false, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": 35, "artificial_analysis_coding_index": 37.8, "artificial_analysis_math_index": 55.7, "mmlu_pro": 0.882, "gpqa": 0.812, "hle": 0.141, "livecodebench": 0.797, "scicode": 0.499, "math_500": null, "aime": null, "aime_25": 0.557, "ifbench": 0.551, "lcr": 0.48, "terminalbench_hard": 0.318, "tau2": 0.433 } }, { "id": "gemini-3.1-pro-preview", "name": "Google: Gemini 3.1 Pro (Preview)", "short_name": "Gemini 3.1 Pro (Preview)", "description": "Gemini 3.1 Pro preview is built for tasks where simple answers are not enough. Stronger core reasoning for complex coding, math, and long-context workflows, with multimodal support and a reported 77.1% verified score on ARC-AGI-2.", "context_length": 1000000, "created": 1771509627, "owned_by": "google", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 1.9 }, "output": { "unit": "1M tokens", "price_per_million_t": 11.4 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0.2 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": true, "function_call": true, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": 57.2, "artificial_analysis_coding_index": 55.5, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": 0.941, "hle": 0.447, "livecodebench": null, "scicode": 0.589, "math_500": null, "aime": null, "aime_25": null, "ifbench": 0.771, "lcr": 0.727, "terminalbench_hard": 0.538, "tau2": 0.956 }, "discount": { "percentage": 5, "note": "A 5% discount applies to input and output token pricing only." } }, { "id": "claude-sonnet-4.6", "name": "Anthropic: Claude Sonnet 4.6", "short_name": "Claude Sonnet 4.6", "description": "Claude Sonnet 4.6 is Anthropic's most capable Sonnet yet — a full upgrade across coding, computer use, long‑context reasoning, agent planning, and design.", "context_length": 1000000, "created": 1771342990, "owned_by": "anthropic", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 2.85 }, "output": { "unit": "1M tokens", "price_per_million_t": 14.25 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0.3 }, "write": { "5m": { "unit": "1M tokens", "price_per_million_t": 3.74 }, "1h": { "unit": "1M tokens", "price_per_million_t": 5.98 } } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": true, "function_call": true, "reasoning": false }, "benchmarks": { "artificial_analysis_intelligence_index": 51.7, "artificial_analysis_coding_index": 50.9, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": 0.875, "hle": 0.3, "livecodebench": null, "scicode": 0.468, "math_500": null, "aime": null, "aime_25": null, "ifbench": 0.566, "lcr": 0.707, "terminalbench_hard": 0.53, "tau2": 0.757 }, "discount": { "percentage": 5, "note": "A 5% discount applies to input and output token pricing only." } }, { "id": "grok-4.1-fast", "name": "xAI: Grok 4.1 Fast", "short_name": "Grok 4.1 Fast", "description": "Grok 4.1 Fast is xAI’s best agentic tool-calling model, optimized for high-performance multimodal workflows. It provides a 2M-token context window and excels at complex real-world tasks such as customer support, finance, and deep research.", "context_length": 2000000, "created": 1763587502, "owned_by": "xai", "available": true, "hallucination_rate": 4.22, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.19 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.48 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0.05 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": true, "function_call": true, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": 23.6, "artificial_analysis_coding_index": 19.5, "artificial_analysis_math_index": 34.3, "mmlu_pro": 0.743, "gpqa": 0.637, "hle": 0.05, "livecodebench": 0.399, "scicode": 0.296, "math_500": null, "aime": null, "aime_25": 0.343, "ifbench": 0.365, "lcr": 0.22, "terminalbench_hard": 0.144, "tau2": 0.637 }, "outdated": true }, { "id": "grok-4-fast", "name": "xAI: Grok 4 Fast", "short_name": "Grok 4 Fast", "description": "Grok 4 Fast, xAI’s latest advancement in cost‑efficient reasoning. Built on learnings from Grok 4, it blends reasoning and non‑reasoning in one model with a 2M‑token context window and state‑of‑the‑art cost efficiency.", "context_length": 2000000, "created": 1758240090, "owned_by": "xai", "available": true, "hallucination_rate": 12.09, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.19 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.48 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0.05 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": true, "function_call": true, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": 23.1, "artificial_analysis_coding_index": 19, "artificial_analysis_math_index": 41.3, "mmlu_pro": 0.73, "gpqa": 0.606, "hle": 0.05, "livecodebench": 0.401, "scicode": 0.329, "math_500": null, "aime": null, "aime_25": 0.413, "ifbench": 0.377, "lcr": 0.2, "terminalbench_hard": 0.121, "tau2": 0.637 }, "outdated": true }, { "id": "deepseek-r1", "name": "DeepSeek: R1", "short_name": "DeepSeek R1", "description": "DeepSeek-R1 is the first-generation reasoning model built atop DeepSeek-V3 (671B total parameters, 37B activated per token). It incorporates large-scale reinforcement learning (RL) to enhance its chain-of-thought and reasoning capabilities, delivering strong performance in math, code, and multi-step reasoning tasks.", "context_length": 163840, "created": 1737381095, "owned_by": "deepseek-ai", "available": true, "hallucination_rate": 14.3, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.37 }, "output": { "unit": "1M tokens", "price_per_million_t": 1.56 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": false, "function_call": true, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": 18.8, "artificial_analysis_coding_index": 15.9, "artificial_analysis_math_index": 68, "mmlu_pro": 0.844, "gpqa": 0.708, "hle": 0.093, "livecodebench": 0.617, "scicode": 0.357, "math_500": 0.966, "aime": 0.683, "aime_25": 0.68, "ifbench": 0.39, "lcr": 0.523, "terminalbench_hard": 0.061, "tau2": 0.114 } }, { "id": "grok-4-07-09", "name": "xAI: Grok 4", "short_name": "Grok 4", "description": "Grok 4 0709 by xAI. Their latest and greatest flagship model, offering unparalleled performance in natural language, math and reasoning - the perfect jack of all trades.", "context_length": 256000, "created": 1752087689, "owned_by": "xai", "available": true, "hallucination_rate": 12.09, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 2.85 }, "output": { "unit": "1M tokens", "price_per_million_t": 14.25 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0.75 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": true, "function_call": true, "web_search": false, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": 41.5, "artificial_analysis_coding_index": 40.5, "artificial_analysis_math_index": 92.7, "mmlu_pro": 0.866, "gpqa": 0.877, "hle": 0.239, "livecodebench": 0.819, "scicode": 0.457, "math_500": 0.99, "aime": 0.943, "aime_25": 0.927, "ifbench": 0.537, "lcr": 0.68, "terminalbench_hard": 0.379, "tau2": 0.749 }, "outdated": true }, { "id": "grok-code-fast-1", "name": "xAI: Grok Code Fast 1", "short_name": "Grok Code Fast 1", "description": "Fast, cost-efficient reasoning model built specifically for agentic coding, trained on programming-rich data and optimized for rapid tool use, real-world coding tasks, and responsive performance across languages like Python, TypeScript, and C++.", "context_length": 256000, "created": 1756238927, "owned_by": "xai", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.19 }, "output": { "unit": "1M tokens", "price_per_million_t": 1.43 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0.02 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": false, "function_call": true, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": 28.7, "artificial_analysis_coding_index": 23.7, "artificial_analysis_math_index": 43.3, "mmlu_pro": 0.793, "gpqa": 0.727, "hle": 0.075, "livecodebench": 0.657, "scicode": 0.362, "math_500": null, "aime": null, "aime_25": 0.433, "ifbench": 0.414, "lcr": 0.483, "terminalbench_hard": 0.174, "tau2": 0.757 } }, { "id": "phoenix-1.0", "name": "Leonardo: Phoenix 1.0", "short_name": "Phoenix 1.0", "description": "Leonardo.AI's state-of-the-art model designed for incredible prompt adherence and crystal-clear text rendition, perfect for creating highly specific and coherent visuals.", "created": 1738368000, "owned_by": "leonardo", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/images/generations", "/v1/images/edits" ], "pricing": { "unit": "per image", "resolutions": { "1792x1024": 0.012, "1024x1792": 0.012, "1024x768": 0.012, "768x1024": 0.012, "1080x1350": 0.012, "auto": 0.012 } }, "benchmarks": { "artificial_analysis_intelligence_index": 10.4, "artificial_analysis_coding_index": 11.2, "artificial_analysis_math_index": 18, "mmlu_pro": 0.714, "gpqa": 0.575, "hle": 0.041, "livecodebench": 0.231, "scicode": 0.26, "math_500": 0.81, "aime": 0.143, "aime_25": 0.18, "ifbench": 0.235, "lcr": 0, "terminalbench_hard": 0.038, "tau2": 0, "elo": 982, "rank": 94, "ci95": "-12/12", "appearances": 3528, "categories": { "style_general_photorealistic": { "elo": 959, "ci95": "-19/19", "appearances": 1485 }, "style_cartoon_illustration": { "elo": 1046, "ci95": "-44/44", "appearances": 263 }, "subject_people:_portraits": { "elo": 954, "ci95": "-44/44", "appearances": 279 }, "style_anime": { "elo": 970, "ci95": "-47/47", "appearances": 242 }, "subject_futuristic_sci-fi": { "elo": 990, "ci95": "-32/32", "appearances": 480 }, "subject_physical_spaces": { "elo": 932, "ci95": "-31/31", "appearances": 563 }, "subject_text_typography": { "elo": 979, "ci95": "-60/60", "appearances": 146 }, "subject_nature_landscapes": { "elo": 1008, "ci95": "-35/35", "appearances": 406 }, "subject_fantasy_mythical": { "elo": 1006, "ci95": "-31/31", "appearances": 539 }, "subject_commercial": { "elo": 1022, "ci95": "-52/52", "appearances": 195 }, "style_traditional_art": { "elo": 1063, "ci95": "-51/51", "appearances": 196 }, "style_graphic_design_digital_rendering": { "elo": 968, "ci95": "-22/22", "appearances": 1024 }, "subject_people:_groups_activities": { "elo": 1008, "ci95": "-42/42", "appearances": 312 } } }, "supported_sizes": [ "1792x1024", "1024x1792", "1024x768", "768x1024", "1080x1350" ] }, { "id": "lucid-origin", "name": "Leonardo: Lucid Origin", "short_name": "Lucid Origin", "description": "Leonardo.AI's most versatile and responsive model, built to handle everything from sharp graphic design and HD renders to complex creative directions. It excels at following detailed prompts, rendering accurate text, and mastering styles ranging from stylized concept art to realistic product mockups.", "created": 1738368000, "owned_by": "leonardo", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/images/generations", "/v1/images/edits" ], "pricing": { "unit": "per image", "resolutions": { "1:1": 0.033, "16:9": 0.033, "9:16": 0.033, "4:3": 0.033, "3:4": 0.033, "2:3": 0.033, "3:2": 0.033, "4:5": 0.033, "5:4": 0.033, "2:1": 0.033, "1:2": 0.033, "3:1": 0.033, "1:3": 0.033, "auto": 0.033 } }, "supported_sizes": [ "16:9", "9:16", "1:1", "4:3", "3:4", "3:2", "2:3", "4:5", "5:4", "2:1", "1:2", "3:1", "1:3" ], "benchmarks": { "elo": 1106, "rank": 38, "ci95": "-11/11", "appearances": 3598, "categories": { "style_general_photorealistic": { "elo": 1132, "ci95": "-18/18", "appearances": 1468 }, "style_cartoon_illustration": { "elo": 1105, "ci95": "-39/39", "appearances": 327 }, "subject_people:_portraits": { "elo": 1053, "ci95": "-34/34", "appearances": 425 }, "style_anime": { "elo": 1120, "ci95": "-44/44", "appearances": 255 }, "subject_futuristic_sci-fi": { "elo": 1098, "ci95": "-34/34", "appearances": 418 }, "subject_physical_spaces": { "elo": 1107, "ci95": "-32/32", "appearances": 468 }, "subject_nature_landscapes": { "elo": 1132, "ci95": "-35/35", "appearances": 381 }, "style_vintage_retro": { "elo": 1156, "ci95": "-58/58", "appearances": 153 }, "subject_fantasy_mythical": { "elo": 1122, "ci95": "-30/30", "appearances": 519 }, "subject_commercial": { "elo": 1105, "ci95": "-47/47", "appearances": 216 }, "style_traditional_art": { "elo": 1162, "ci95": "-46/46", "appearances": 232 }, "style_graphic_design_digital_rendering": { "elo": 1057, "ci95": "-22/22", "appearances": 978 }, "subject_people:_groups_activities": { "elo": 1179, "ci95": "-37/37", "appearances": 363 } } } }, { "id": "flux-2-dev", "name": "Black Forest Labs: FLUX.2 [dev]", "short_name": "FLUX.2 [dev]", "description": "FLUX.2 [dev] is an image model from Black Forest Labs where you can generate highly realistic and detailed images, with multi-reference support.", "created": 1738368000, "owned_by": "blackforestlabs", "available": true, "hallucination_rate": 0, "supported_aspect_ratios": [], "endpoints": [ "/v1/images/generations", "/v1/images/edits" ], "pricing": { "unit": "per image", "resolutions": { "1024x1024": 0.02244, "1280x720": 0.02244, "720x1280": 0.02244, "1536x1024": 0.02244, "1024x1536": 0.02244, "auto": 0.02244 } }, "supported_sizes": [ "1024x1024", "1280x720", "720x1280", "1536x1024", "1024x1536" ], "benchmarks": null }, { "id": "flux-1-schnell", "name": "Black Forest Labs: FLUX.1 [schnell]", "short_name": "FLUX.1 [schnell]", "description": "A lightning-fast text-to-image model from Black Forest Labs, optimized for rapid generation and scalable creative workflows.", "created": 1738368000, "owned_by": "blackforestlabs", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/images/generations", "/v1/images/edits" ], "pricing": { "unit": "per image", "resolutions": { "1792x1024": 0.0014, "1024x1792": 0.0014, "1024x768": 0.0014, "768x1024": 0.0014, "1080x1350": 0.0014, "auto": 0.0014 } }, "benchmarks": null, "supported_sizes": [ "1792x1024", "1024x1792", "1024x768", "768x1024", "1080x1350" ] }, { "id": "sdxl", "name": "Stability: Stable Diffusion XL", "short_name": "SDXL", "description": "Stable Diffusion XL (SDXL) is a high-quality AI image generator that creates detailed images from text prompts with improved realism and control.", "created": 1738368000, "owned_by": "stabilityai", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/images/generations", "/v1/images/edits" ], "pricing": { "unit": "per image", "resolutions": { "512x512": 0.00935, "768x768": 0.00935, "1024x1024": 0.00935, "1408x1408": 0.00935, "576x1024": 0.00935, "1024x576": 0.00935, "768x1024": 0.00935, "1024x768": 0.00935, "auto": 0.00935 } }, "benchmarks": { "artificial_analysis_intelligence_index": 18.8, "artificial_analysis_coding_index": 13.6, "artificial_analysis_math_index": 30.3, "mmlu_pro": 0.76, "gpqa": 0.578, "hle": 0.043, "livecodebench": 0.4, "scicode": 0.331, "math_500": 0.907, "aime": 0.44, "aime_25": 0.303, "ifbench": 0.393, "lcr": 0.28, "terminalbench_hard": 0.038, "tau2": 0.243 }, "supported_sizes": [ "512x512", "768x768", "1024x1024", "1408x1408", "576x1024", "1024x576", "768x1024", "1024x768" ] }, { "id": "flux-2-klein-9b", "name": "Black Forest Labs: FLUX.2 [klein] 9B", "short_name": "FLUX.2 [klein] 9B", "description": "A highly optimized, distilled version of the FLUX architecture designed for lightning-fast generations without compromising on quality. It specializes in real-time creative workflows, offering an ideal balance of speed and artistic detail for high-performance applications.", "created": 1738368000, "owned_by": "blackforestlabs", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/images/generations", "/v1/images/edits" ], "pricing": { "unit": "per image", "resolutions": { "1024x1024": 0.01122, "1280x720": 0.01122, "720x1280": 0.01122, "1536x1024": 0.01122, "1024x1536": 0.01122, "1536x1536": 0.01122, "auto": 0.01122 } }, "supported_sizes": [ "1024x1024", "1280x720", "720x1280", "1536x1024", "1024x1536", "1536x1536" ], "benchmarks": { "elo": 1135, "rank": 24, "ci95": "-10/10", "appearances": 4590, "categories": { "style_general_photorealistic": { "elo": 1167, "ci95": "-16/16", "appearances": 1811 }, "style_cartoon_illustration": { "elo": 1139, "ci95": "-35/35", "appearances": 428 }, "subject_people:_portraits": { "elo": 1129, "ci95": "-29/29", "appearances": 571 }, "style_anime": { "elo": 1145, "ci95": "-36/36", "appearances": 393 }, "subject_futuristic_sci-fi": { "elo": 1137, "ci95": "-28/28", "appearances": 640 }, "subject_physical_spaces": { "elo": 1154, "ci95": "-28/28", "appearances": 634 }, "subject_text_typography": { "elo": 1049, "ci95": "-54/54", "appearances": 169 }, "subject_nature_landscapes": { "elo": 1122, "ci95": "-32/32", "appearances": 452 }, "style_vintage_retro": { "elo": 1109, "ci95": "-49/49", "appearances": 211 }, "subject_fantasy_mythical": { "elo": 1177, "ci95": "-26/26", "appearances": 719 }, "subject_ui/ux_design": { "elo": 1024, "ci95": "-64/64", "appearances": 131 }, "subject_commercial": { "elo": 1133, "ci95": "-44/44", "appearances": 261 }, "style_traditional_art": { "elo": 1221, "ci95": "-39/39", "appearances": 332 }, "style_graphic_design_digital_rendering": { "elo": 1076, "ci95": "-20/20", "appearances": 1223 }, "subject_people:_groups_activities": { "elo": 1204, "ci95": "-33/33", "appearances": 484 } } } }, { "id": "flux-2-klein-4b", "name": "Black Forest Labs: FLUX.2 [klein] 4B", "short_name": "FLUX.2 [klein] 4B", "description": "A highly optimized, distilled version of the FLUX architecture designed for lightning-fast generations without compromising on quality. It specializes in real-time creative workflows, offering an ideal balance of speed and artistic detail for high-performance applications.", "created": 1738368000, "owned_by": "blackforestlabs", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/images/generations", "/v1/images/edits" ], "pricing": { "unit": "per image", "resolutions": { "1024x1024": 0.01122, "1280x720": 0.01122, "720x1280": 0.01122, "1536x1024": 0.01122, "1024x1536": 0.01122, "1536x1536": 0.01122, "auto": 0.01122 } }, "supported_sizes": [ "1024x1024", "1280x720", "720x1280", "1536x1024", "1024x1536", "1536x1536" ], "benchmarks": { "elo": 1069, "rank": 54, "ci95": "-10/10", "appearances": 4451, "categories": { "style_general_photorealistic": { "elo": 1105, "ci95": "-16/16", "appearances": 1774 }, "style_cartoon_illustration": { "elo": 1066, "ci95": "-36/36", "appearances": 392 }, "subject_people:_portraits": { "elo": 1090, "ci95": "-30/30", "appearances": 538 }, "style_anime": { "elo": 1086, "ci95": "-37/37", "appearances": 355 }, "subject_futuristic_sci-fi": { "elo": 1060, "ci95": "-28/28", "appearances": 598 }, "subject_physical_spaces": { "elo": 1089, "ci95": "-28/28", "appearances": 602 }, "subject_text_typography": { "elo": 944, "ci95": "-55/55", "appearances": 183 }, "subject_nature_landscapes": { "elo": 1116, "ci95": "-33/33", "appearances": 445 }, "style_vintage_retro": { "elo": 983, "ci95": "-52/52", "appearances": 202 }, "subject_fantasy_mythical": { "elo": 1122, "ci95": "-25/25", "appearances": 749 }, "subject_ui/ux_design": { "elo": 949, "ci95": "-64/64", "appearances": 133 }, "subject_commercial": { "elo": 1071, "ci95": "-47/47", "appearances": 225 }, "style_traditional_art": { "elo": 1176, "ci95": "-40/40", "appearances": 304 }, "style_graphic_design_digital_rendering": { "elo": 1015, "ci95": "-20/20", "appearances": 1221 }, "subject_people:_groups_activities": { "elo": 1108, "ci95": "-33/33", "appearances": 471 } } } }, { "id": "kimi-k2.5", "name": "MoonshotAI: Kimi K2.5", "short_name": "Kimi K2.5", "description": "Kimi K2.5 is Moonshot AI's native multimodal model built on Kimi K2 with ~15T mixed visual and text tokens, delivering strong general reasoning, visual coding, and agentic tool-calling.", "context_length": 262144, "created": 1769527976, "owned_by": "moonshotai", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.34 }, "output": { "unit": "1M tokens", "price_per_million_t": 2.17 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": true, "function_call": true, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": 46.8, "artificial_analysis_coding_index": 39.5, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": 0.879, "hle": 0.294, "livecodebench": null, "scicode": 0.49, "math_500": null, "aime": null, "aime_25": null, "ifbench": 0.702, "lcr": 0.653, "terminalbench_hard": 0.348, "tau2": 0.959 }, "sub": true }, { "id": "minimax-m2.5", "name": "MiniMax: MiniMax M2.5", "short_name": "MiniMax M2.5", "description": "MiniMax M2.5 is a productivity-focused flagship model that builds on M2.1 with stronger coding and real-world office workflow performance (Word, Excel, PowerPoint), plus better tool-use planning and token efficiency.", "context_length": 204800, "created": 1770910378, "owned_by": "minimax", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.34 }, "output": { "unit": "1M tokens", "price_per_million_t": 1.37 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0.055 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": false, "function_call": true, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": 41.9, "artificial_analysis_coding_index": 37.4, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": 0.848, "hle": 0.191, "livecodebench": null, "scicode": 0.426, "math_500": null, "aime": null, "aime_25": null, "ifbench": 0.716, "lcr": 0.66, "terminalbench_hard": 0.348, "tau2": 0.953 }, "sub": true }, { "id": "claude-opus-4-6", "name": "Anthropic: Claude Opus 4.6", "short_name": "Claude Opus 4.6", "description": "Opus 4.6 is Anthropic’s most capable model for coding and long-running work. Built for full workflows, it excels at large codebases, complex refactors, and multi-step debugging, with stronger context and reliability than earlier versions. It also performs well on sustained knowledge work, producing near-production-ready documents and plans in a single pass and staying coherent over long sessions. See the migration guide for upgrading from earlier Opus versions.", "context_length": 1000000, "created": 1770219050, "owned_by": "anthropic", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 4.5 }, "output": { "unit": "1M tokens", "price_per_million_t": 22.5 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0.5 }, "write": { "5m": { "unit": "1M tokens", "price_per_million_t": 6.25 }, "1h": { "unit": "1M tokens", "price_per_million_t": 10 } } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": true, "function_call": true, "reasoning": false }, "benchmarks": { "artificial_analysis_intelligence_index": 53, "artificial_analysis_coding_index": 48.1, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": 0.896, "hle": 0.367, "livecodebench": null, "scicode": 0.519, "math_500": null, "aime": null, "aime_25": null, "ifbench": 0.531, "lcr": 0.707, "terminalbench_hard": 0.462, "tau2": 0.921 }, "discount": { "percentage": 5, "note": "A 5% discount applies to input and output token pricing only." } }, { "id": "glm-5", "name": "Z.ai: GLM 5", "short_name": "GLM 5", "description": "GLM-5 is Z.ai’s flagship open-source foundation model engineered for complex systems design and long-horizon agent workflows. Built for expert developers, it delivers production-grade performance on large-scale programming tasks, rivaling leading closed-source models. With advanced agentic planning, deep backend reasoning, and iterative self-correction, GLM-5 moves beyond code generation to full-system construction and autonomous execution.", "context_length": 200000, "created": 1770829182, "owned_by": "z-ai", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.3 }, "output": { "unit": "1M tokens", "price_per_million_t": 2.55 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0.23 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": false, "function_call": true, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": 49.8, "artificial_analysis_coding_index": 44.2, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": 0.82, "hle": 0.272, "livecodebench": null, "scicode": 0.462, "math_500": null, "aime": null, "aime_25": null, "ifbench": 0.723, "lcr": 0.633, "terminalbench_hard": 0.432, "tau2": 0.982 } }, { "id": "gemma-3n-e2b-it", "name": "Google: Gemma 3n 2B", "short_name": "Gemma 3n 2B", "description": "Gemma 3n E2B IT is a multimodal, instruction-tuned model developed by Google DeepMind, designed to operate efficiently at an effective parameter size of 2B while leveraging a 6B architecture. Based on the MatFormer architecture, it supports nested submodels and modular composition via the Mix-and-Match framework. Gemma 3n models are optimized for low-resource deployment, offering 32K context length and strong multilingual and reasoning performance across common benchmarks. This variant is trained on a diverse corpus including code, math, web, and multimodal data.", "context_length": 8192, "created": 1752074904, "owned_by": "google", "available": true, "hallucination_rate": 0.9262, "endpoints": [ "/v1/chat/completions" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.03 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.02 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": false, "function_call": true, "reasoning": false }, "benchmarks": { "artificial_analysis_intelligence_index": 7.7, "artificial_analysis_coding_index": 0, "artificial_analysis_math_index": 2.3, "mmlu_pro": 0.055, "gpqa": 0.224, "hle": 0.042, "livecodebench": 0.003, "scicode": 0, "math_500": 0.691, "aime": 0.09, "aime_25": 0.023, "ifbench": 0.121, "lcr": 0, "terminalbench_hard": 0, "tau2": 0.091 }, "discount": { "percentage": 5, "note": "A 5% discount applies to input and output token pricing only." } }, { "id": "gemma-3n-e4b-it", "name": "Google: Gemma 3n 4B", "short_name": "Gemma 3n 4B", "description": "Gemma 3n E4B-it is optimized for efficient execution on mobile and low-resource devices, such as phones, laptops, and tablets. It supports multimodal inputs—including text, visual data, and audio—enabling diverse tasks such as text generation, speech recognition, translation, and image analysis. Leveraging innovations like Per-Layer Embedding (PLE) caching and the MatFormer architecture, Gemma 3n dynamically manages memory usage and computational load by selectively activating model parameters, significantly reducing runtime resource requirements. This model supports a wide linguistic range (trained in over 140 languages) and features a flexible 32K token context window. Gemma 3n can selectively load parameters, optimizing memory and computational efficiency based on the task or device capabilities, making it well-suited for privacy-focused, offline-capable applications and on-device AI solutions. [Read more in the blog post](https://developers.googleblog.com/en/introducing-gemma-3n/)", "context_length": 8192, "created": 1747776824, "owned_by": "google", "available": true, "hallucination_rate": 0.9656, "endpoints": [ "/v1/chat/completions" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.05 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.04 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": false, "function_call": true, "reasoning": false }, "benchmarks": { "artificial_analysis_intelligence_index": 6.4, "artificial_analysis_coding_index": 4.2, "artificial_analysis_math_index": 14.3, "mmlu_pro": 0.488, "gpqa": 0.296, "hle": 0.044, "livecodebench": 0.146, "scicode": 0.081, "math_500": 0.771, "aime": 0.137, "aime_25": 0.143, "ifbench": 0.279, "lcr": 0, "terminalbench_hard": 0.023, "tau2": 0.05 }, "discount": { "percentage": 5, "note": "A 5% discount applies to input and output token pricing only." } }, { "id": "gemma-3-4b-it", "name": "Google: Gemma 3 4B", "short_name": "Gemma 3 4B", "description": "Gemma 3 introduces multimodality, supporting vision-language input and text outputs. It understands over 140 languages, and offers improved math, reasoning, and chat capabilities, including structured outputs and function calling.", "context_length": 32768, "created": 1741905510, "owned_by": "google", "available": true, "hallucination_rate": 6.4, "endpoints": [ "/v1/chat/completions" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.01 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.03 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": false, "function_call": true, "reasoning": false }, "benchmarks": { "artificial_analysis_intelligence_index": 6.3, "artificial_analysis_coding_index": 2.9, "artificial_analysis_math_index": 12.7, "mmlu_pro": 0.417, "gpqa": 0.291, "hle": 0.052, "livecodebench": 0.112, "scicode": 0.073, "math_500": 0.766, "aime": 0.063, "aime_25": 0.127, "ifbench": 0.283, "lcr": 0.057, "terminalbench_hard": 0.008, "tau2": 0.05 }, "discount": { "percentage": 5, "note": "A 5% discount applies to input and output token pricing only." } }, { "id": "gemma-3-12b-it", "name": "Google: Gemma 3 12B", "short_name": "Gemma 3 12B", "description": "Gemma 3 introduces multimodality, supporting vision-language input and text outputs. It understands over 140 languages, and offers improved math, reasoning, and chat capabilities, including structured outputs and function calling. Gemma 3 12B is the second largest in the family of Gemma 3 models after Gemma 3 27B.", "context_length": 32768, "created": 1741902625, "owned_by": "google", "available": true, "hallucination_rate": 4.4, "endpoints": [ "/v1/chat/completions" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.03 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.1 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": false, "function_call": true, "reasoning": false }, "benchmarks": { "artificial_analysis_intelligence_index": 7.7, "artificial_analysis_coding_index": 0, "artificial_analysis_math_index": 2.3, "mmlu_pro": 0.055, "gpqa": 0.224, "hle": 0.042, "livecodebench": 0.003, "scicode": 0, "math_500": 0.853, "aime": 0.22, "aime_25": 0.023, "ifbench": 0.121, "lcr": 0, "terminalbench_hard": 0, "tau2": 0.091 }, "discount": { "percentage": 5, "note": "A 5% discount applies to input and output token pricing only." } }, { "id": "gemma-3-27b-it", "name": "Google: Gemma 3 27B", "short_name": "Gemma 3 27B", "description": "Gemma 3 introduces multimodality, supporting vision-language input and text outputs. It handles context windows up to 128k tokens, understands over 140 languages, and offers improved math, reasoning, and chat capabilities, including structured outputs and function calling.", "context_length": 131072, "created": 1741756359, "owned_by": "google", "available": true, "hallucination_rate": 7.4, "endpoints": [ "/v1/chat/completions" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.04 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.15 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": true, "function_call": true, "reasoning": false }, "benchmarks": { "artificial_analysis_intelligence_index": 7.7, "artificial_analysis_coding_index": 0, "artificial_analysis_math_index": 2.3, "mmlu_pro": 0.055, "gpqa": 0.224, "hle": 0.042, "livecodebench": 0.003, "scicode": 0, "math_500": 0.883, "aime": 0.253, "aime_25": 0.023, "ifbench": 0.121, "lcr": 0, "terminalbench_hard": 0, "tau2": 0.091 }, "discount": { "percentage": 5, "note": "A 5% discount applies to input and output token pricing only." } }, { "id": "command-a-03-2025", "name": "Cohere: Command A (03-2025)", "short_name": "Command A (03-2025)", "description": "Command A is Cohere's most performant model to date, excelling at tool use, agents, retrieval augmented generation (RAG), and multilingual use cases. Command A has a context length of 256K, only requires two GPUs to run, and has 150% higher throughput compared to Command R+ 08-2024.", "context_length": 256000, "created": 1738368000, "owned_by": "cohere", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 2.2 }, "output": { "unit": "1M tokens", "price_per_million_t": 10 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": false, "function_call": true, "reasoning": false }, "benchmarks": { "artificial_analysis_intelligence_index": 13.5, "artificial_analysis_coding_index": 9.9, "artificial_analysis_math_index": 13, "mmlu_pro": 0.712, "gpqa": 0.527, "hle": 0.046, "livecodebench": 0.287, "scicode": 0.281, "math_500": 0.819, "aime": 0.097, "aime_25": 0.13, "ifbench": 0.365, "lcr": 0.18, "terminalbench_hard": 0.008, "tau2": 0.152 } }, { "id": "command-r7b-12-2024", "name": "Cohere: Command R7B (12-2024)", "short_name": "Command R7B (12-2024)", "description": "command-r7b-12-2024 is a small, fast update delivered in December 2024. It excels at RAG, tool use, agents, and similar tasks requiring complex reasoning and multiple steps.", "context_length": 128000, "created": 1733011200, "owned_by": "cohere", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.0375 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.13 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": false, "function_call": true, "reasoning": false }, "benchmarks": { "artificial_analysis_intelligence_index": 7.4, "artificial_analysis_coding_index": null, "artificial_analysis_math_index": null, "mmlu_pro": 0.338, "gpqa": 0.284, "hle": 0.048, "livecodebench": 0.048, "scicode": 0.062, "math_500": 0.164, "aime": 0.007, "aime_25": null, "ifbench": null, "lcr": null, "terminalbench_hard": null, "tau2": null } }, { "id": "command-a-translate-08-2025", "name": "Cohere: Command A Translate (08-2025)", "short_name": "Command A Translate", "description": "Command A Translate is Cohere's state of the art machine translation model, excelling at a variety of translation tasks on 23 languages: English, French, Spanish, Italian, German, Portuguese, Japanese, Korean, Chinese, Arabic, Russian, Polish, Turkish, Vietnamese, Dutch, Czech, Indonesian, Ukrainian, Romanian, Greek, Hindi, Hebrew, Persian.", "context_length": 8000, "created": 1754006400, "owned_by": "cohere", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 2.2 }, "output": { "unit": "1M tokens", "price_per_million_t": 10 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": false, "function_call": true, "reasoning": false }, "benchmarks": { "artificial_analysis_intelligence_index": 7.4, "artificial_analysis_coding_index": null, "artificial_analysis_math_index": null, "mmlu_pro": 0.338, "gpqa": 0.284, "hle": 0.048, "livecodebench": 0.048, "scicode": 0.062, "math_500": 0.164, "aime": 0.007, "aime_25": null, "ifbench": null, "lcr": null, "terminalbench_hard": null, "tau2": null } }, { "id": "command-a-reasoning-08-2025", "name": "Cohere: Command A Reasoning (08-2025)", "short_name": "Command A Reasoning", "description": "Command A Reasoning is Cohere's first reasoning model, able to 'think' before generating an output in a way that allows it to perform well in certain kinds of nuanced problem-solving and agent-based tasks in 23 languages.", "context_length": 256000, "created": 1754006400, "owned_by": "cohere", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 2.3 }, "output": { "unit": "1M tokens", "price_per_million_t": 10 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": false, "function_call": true, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": 24.3, "artificial_analysis_coding_index": null, "artificial_analysis_math_index": null, "mmlu_pro": 0.8, "gpqa": 0.698, "hle": 0.116, "livecodebench": 0.505, "scicode": 0.359, "math_500": 0.981, "aime": 0.843, "aime_25": null, "ifbench": null, "lcr": null, "terminalbench_hard": null, "tau2": null } }, { "id": "command-r-08-2024", "name": "Cohere: Command R (08-2024)", "short_name": "Command R (08-2024)", "description": "command-r-08-2024 is an update of the Command R model, delivered in August 2024.", "context_length": 128000, "created": 1723420800, "owned_by": "cohere", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.15 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.6 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": false, "function_call": true, "reasoning": false }, "benchmarks": { "artificial_analysis_intelligence_index": 7.4, "artificial_analysis_coding_index": null, "artificial_analysis_math_index": null, "mmlu_pro": 0.338, "gpqa": 0.284, "hle": 0.048, "livecodebench": 0.048, "scicode": 0.062, "math_500": 0.164, "aime": 0.007, "aime_25": null, "ifbench": null, "lcr": null, "terminalbench_hard": null, "tau2": null } }, { "id": "glm-4.7", "name": "Z.AI: GLM 4.7", "short_name": "GLM 4.7", "description": "GLM-4.7 is Z.AI’s latest flagship model, featuring upgrades in two key areas: enhanced programming capabilities and more stable multi-step reasoning/execution. It demonstrates significant improvements in executing complex agent tasks while delivering more natural conversational experiences and superior front-end aesthetics.", "context_length": 202752, "created": 1766378014, "owned_by": "z-ai", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.15 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.8 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0.1 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": false, "function_call": true, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": 42.1, "artificial_analysis_coding_index": 36.3, "artificial_analysis_math_index": 95, "mmlu_pro": 0.856, "gpqa": 0.859, "hle": 0.251, "livecodebench": 0.894, "scicode": 0.451, "math_500": null, "aime": null, "aime_25": 0.95, "ifbench": 0.679, "lcr": 0.64, "terminalbench_hard": 0.318, "tau2": 0.959 }, "sub": true }, { "id": "nemotron-3-nano-30b-a3b", "name": "NVIDIA: Nemotron 3 Nano 30B A3B", "short_name": "Nemotron 3 Nano 30B A3B", "description": "NVIDIA Nemotron 3 Nano 30B A3B is a small language MoE model with highest compute efficiency and accuracy for developers to build specialized agentic AI systems. The model is fully open with open-weights, datasets and recipes so developers can easily customize, optimize, and deploy the model on their infrastructure for maximum privacy and security. ", "context_length": 256000, "created": 1765731275, "owned_by": "nvidia", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.1 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.15 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0.2 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": false, "function_call": true, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": 13.2, "artificial_analysis_coding_index": 15.8, "artificial_analysis_math_index": 13.3, "mmlu_pro": 0.579, "gpqa": 0.399, "hle": 0.046, "livecodebench": 0.36, "scicode": 0.23, "math_500": null, "aime": null, "aime_25": 0.133, "ifbench": 0.375, "lcr": 0.067, "terminalbench_hard": 0.121, "tau2": 0.254 } }, { "id": "nemotron-3-nano-30b-a3b:free", "name": "NVIDIA: Nemotron 3 Nano 30B A3B (Free)", "short_name": "Nemotron 3 Nano 30B A3B (Free)", "description": "NVIDIA Nemotron 3 Nano 30B A3B is a small language MoE model with highest compute efficiency and accuracy for developers to build specialized agentic AI systems. The model is fully open with open-weights, datasets and recipes so developers can easily customize, optimize, and deploy the model on their infrastructure for maximum privacy and security. ", "context_length": 256000, "created": 1765731275, "owned_by": "nvidia", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0 }, "output": { "unit": "1M tokens", "price_per_million_t": 0 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": false, "function_call": true, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": 13.2, "artificial_analysis_coding_index": 15.8, "artificial_analysis_math_index": 13.3, "mmlu_pro": 0.579, "gpqa": 0.399, "hle": 0.046, "livecodebench": 0.36, "scicode": 0.23, "math_500": null, "aime": null, "aime_25": 0.133, "ifbench": 0.375, "lcr": 0.067, "terminalbench_hard": 0.121, "tau2": 0.254 } }, { "id": "deepseek-v3.2", "name": "DeepSeek: DeepSeek V3.2", "short_name": "DeepSeek V3.2", "description": "DeepSeek-V3.2 is a large language model designed to harmonize high computational efficiency with strong reasoning and agentic tool-use performance. It introduces DeepSeek Sparse Attention (DSA), a fine-grained sparse attention mechanism that reduces training and inference cost while preserving quality in long-context scenarios. A scalable reinforcement learning post-training framework further improves reasoning, with reported performance in the GPT-5 class, and the model has demonstrated gold-medal results on the 2025 IMO and IOI. V3.2 also uses a large-scale agentic task synthesis pipeline to better integrate reasoning into tool-use settings, boosting compliance and generalization in interactive environments.", "context_length": 163840, "created": 1764594642, "owned_by": "deepseek-ai", "available": true, "hallucination_rate": 6.7, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.25 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.39 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": false, "function_call": true, "reasoning": false }, "benchmarks": { "artificial_analysis_intelligence_index": 32.1, "artificial_analysis_coding_index": 34.6, "artificial_analysis_math_index": 59, "mmlu_pro": 0.837, "gpqa": 0.751, "hle": 0.105, "livecodebench": 0.593, "scicode": 0.387, "math_500": null, "aime": null, "aime_25": 0.59, "ifbench": 0.49, "lcr": 0.39, "terminalbench_hard": 0.326, "tau2": 0.789 }, "sub": true }, { "id": "kimi-k2-thinking", "name": "MoonshotAI: Kimi K2 Thinking", "short_name": "Kimi K2 Thinking", "description": "Kimi K2 Thinking is Moonshot AI’s most advanced open reasoning model to date, extending the K2 series into agentic, long-horizon reasoning. Built on the trillion-parameter Mixture-of-Experts (MoE) architecture introduced in Kimi K2, it activates 32 billion parameters per forward pass and supports 256 k-token context windows. The model is optimized for persistent step-by-step thought, dynamic tool invocation, and complex reasoning workflows that span hundreds of turns. It interleaves step-by-step reasoning with tool use, enabling autonomous research, coding, and writing that can persist for hundreds of sequential actions without drift.\n\nIt sets new open-source benchmarks on HLE, BrowseComp, SWE-Multilingual, and LiveCodeBench, while maintaining stable multi-agent behavior through 200–300 tool calls. Built on a large-scale MoE architecture with MuonClip optimization, it combines strong reasoning depth with high inference efficiency for demanding agentic and analytical tasks.", "context_length": 262144, "created": 1762440622, "owned_by": "moonshotai", "available": true, "hallucination_rate": 1.6, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.3 }, "output": { "unit": "1M tokens", "price_per_million_t": 2 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": false, "function_call": true, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": 40.9, "artificial_analysis_coding_index": 34.8, "artificial_analysis_math_index": 94.7, "mmlu_pro": 0.848, "gpqa": 0.838, "hle": 0.223, "livecodebench": 0.853, "scicode": 0.424, "math_500": null, "aime": null, "aime_25": 0.947, "ifbench": 0.681, "lcr": 0.663, "terminalbench_hard": 0.311, "tau2": 0.93 } }, { "id": "devstral-2512", "name": "Mistral: Devstral 2 2512", "short_name": "Devstral 2 2512", "description": "Devstral 2 is a state-of-the-art open-source model by Mistral AI specializing in agentic coding. It is a 123B-parameter dense transformer model supporting a 256K context window. Devstral 2 supports exploring codebases and orchestrating changes across multiple files while maintaining architecture-level context. It tracks framework dependencies, detects failures, and retries with corrections—solving challenges like bug fixing and modernizing legacy systems. The model can be fine-tuned to prioritize specific languages or optimize for large enterprise codebases. It is available under a modified MIT license.", "context_length": 262144, "created": 1765285419, "owned_by": "mistral", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.15 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.6 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": false, "function_call": true, "reasoning": false }, "benchmarks": { "artificial_analysis_intelligence_index": 22, "artificial_analysis_coding_index": 23.7, "artificial_analysis_math_index": 36.7, "mmlu_pro": 0.762, "gpqa": 0.594, "hle": 0.036, "livecodebench": 0.448, "scicode": 0.331, "math_500": null, "aime": null, "aime_25": 0.367, "ifbench": 0.381, "lcr": 0.3, "terminalbench_hard": 0.189, "tau2": 0.249 } }, { "id": "kimi-k2-0905", "name": "MoonshotAI: Kimi K2 0905", "short_name": "Kimi K2 0905", "description": "Kimi K2 0905 is the September update of Kimi K2 0711. It is a large-scale Mixture-of-Experts (MoE) language model developed by Moonshot AI, featuring 1 trillion total parameters with 32 billion active per forward pass. It supports long-context inference up to 256k tokens, extended from the previous 128k.\nThis update improves agentic coding with higher accuracy and better generalization across scaffolds, and enhances frontend coding with more aesthetic and functional outputs for web, 3D, and related tasks. Kimi K2 is optimized for agentic capabilities, including advanced tool use, reasoning, and code synthesis. It excels across coding (LiveCodeBench, SWE-bench), reasoning (ZebraLogic, GPQA), and tool-use (Tau2, AceBench) benchmarks. The model is trained with a novel stack incorporating the MuonClip optimizer for stable large-scale MoE training.", "context_length": 262144, "created": 1757021147, "owned_by": "moonshotai", "available": true, "hallucination_rate": 17.9, "endpoints": [ "/v1/chat/completions" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.39 }, "output": { "unit": "1M tokens", "price_per_million_t": 1.9 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": true, "function_call": true, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": 30.9, "artificial_analysis_coding_index": 25.9, "artificial_analysis_math_index": 57.3, "mmlu_pro": 0.819, "gpqa": 0.767, "hle": 0.063, "livecodebench": 0.61, "scicode": 0.307, "math_500": null, "aime": null, "aime_25": 0.573, "ifbench": 0.417, "lcr": 0.523, "terminalbench_hard": 0.235, "tau2": 0.734 }, "sub": true }, { "id": "laguna-xs.2:free", "name": "Poolside: Laguna XS.2 (Free)", "short_name": "Laguna XS.2 (Free)", "description": "Designed for agentic coding workflows.", "context_length": 131072, "created": 1777388504, "owned_by": "poolside", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0, "price_per_token_usd": "0.0" }, "output": { "unit": "1M tokens", "price_per_million_t": 0, "price_per_token_usd": "0.0" } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "stop", "temperature", "top_p", "tools", "tool_choice" ], "capabilities": { "vision": false, "function_call": true, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": null, "artificial_analysis_coding_index": null, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": null, "hle": null, "livecodebench": null, "scicode": null, "math_500": null, "aime": null, "aime_25": null, "ifbench": null, "lcr": null, "terminalbench_hard": null, "tau2": null } }, { "id": "laguna-m.1:free", "name": "Poolside: Laguna M.1 (Free)", "short_name": "Laguna M.1 (Free)", "description": "Designed for agentic coding workflows", "context_length": 131072, "created": 1777388504, "owned_by": "poolside", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0, "price_per_token_usd": "0.0" }, "output": { "unit": "1M tokens", "price_per_million_t": 0, "price_per_token_usd": "0.0" } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "stop", "temperature", "top_p", "tools", "tool_choice" ], "capabilities": { "vision": false, "function_call": true, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": null, "artificial_analysis_coding_index": null, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": null, "hle": null, "livecodebench": null, "scicode": null, "math_500": null, "aime": null, "aime_25": null, "ifbench": null, "lcr": null, "terminalbench_hard": null, "tau2": null } }, { "id": "minimax-m2", "name": "MiniMax: MiniMax M2", "short_name": "MiniMax M2", "description": "MiniMax-M2 is a compact, fast, and cost-effective Mixture-of-Experts (MoE) model with 230 billion total parameters and 10 billion active parameters, built for elite performance in coding and agentic tasks while maintaining powerful general intelligence.", "context_length": 196608, "created": 1761252093, "owned_by": "minimax", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.25 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.85 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": false, "function_call": true, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": 36.1, "artificial_analysis_coding_index": 29.2, "artificial_analysis_math_index": 78.3, "mmlu_pro": 0.82, "gpqa": 0.777, "hle": 0.125, "livecodebench": 0.826, "scicode": 0.361, "math_500": null, "aime": null, "aime_25": 0.783, "ifbench": 0.723, "lcr": 0.61, "terminalbench_hard": 0.258, "tau2": 0.868 } }, { "id": "nemotron-nano-12b-v2-vl", "name": "NVIDIA: Nemotron Nano 12B 2 VL", "short_name": "Nemotron Nano 12B 2 VL", "description": "NVIDIA Nemotron Nano 2 VL is a 12-billion-parameter open multimodal reasoning model designed for video understanding and document intelligence. It introduces a hybrid Transformer-Mamba architecture, combining transformer-level accuracy with Mamba’s memory-efficient sequence modeling for significantly higher throughput and lower latency. The model supports inputs of text and multi-image documents, producing natural-language outputs. It is trained on high-quality NVIDIA-curated synthetic datasets optimized for optical-character recognition, chart reasoning, and multimodal comprehension. Nemotron Nano 2 VL achieves leading results on OCRBench v2 and scores ≈ 74 average across MMMU, MathVista, AI2D, OCRBench, OCR-Reasoning, ChartQA, DocVQA, and Video-MME—surpassing prior open VL baselines. With Efficient Video Sampling (EVS), it handles long-form videos while reducing inference cost. Open-weights, training data, and fine-tuning recipes are released under a permissive NVIDIA open license, with deployment supported across NeMo, NIM, and major inference runtimes", "context_length": 128000, "created": 1761675565, "owned_by": "nvidia", "available": true, "hallucination_rate": 2.5, "endpoints": [ "/v1/chat/completions" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.02 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.06 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": true, "function_call": true, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": 10.1, "artificial_analysis_coding_index": 5.9, "artificial_analysis_math_index": 26.7, "mmlu_pro": 0.649, "gpqa": 0.439, "hle": 0.045, "livecodebench": 0.345, "scicode": 0.176, "math_500": null, "aime": null, "aime_25": 0.267, "ifbench": 0.259, "lcr": 0.17, "terminalbench_hard": 0, "tau2": 0.193 } }, { "id": "glm-4.6", "name": "Z.AI: GLM 4.6", "short_name": "GLM 4.6", "description": "Compared with GLM-4.5, this generation brings several key improvements:\n\nLonger context window: The context window has been expanded from 128K to 200K tokens, enabling the model to handle more complex agentic tasks.\nSuperior coding performance: The model achieves higher scores on code benchmarks and demonstrates better real-world performance in applications such as Claude Code、Cline、Roo Code and Kilo Code, including improvements in generating visually polished front-end pages.\nAdvanced reasoning: GLM-4.6 shows a clear improvement in reasoning performance and supports tool use during inference, leading to stronger overall capability.\nMore capable agents: GLM-4.6 exhibits stronger performance in tool using and search-based agents, and integrates more effectively within agent frameworks.\nRefined writing: Better aligns with human preferences in style and readability, and performs more naturally in role-playing scenarios.", "context_length": 202752, "created": 1759235576, "owned_by": "z-ai", "available": true, "hallucination_rate": 1.3, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.43 }, "output": { "unit": "1M tokens", "price_per_million_t": 1.75 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0.12 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": false, "function_call": true, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": 30.2, "artificial_analysis_coding_index": 30.2, "artificial_analysis_math_index": 44.3, "mmlu_pro": 0.784, "gpqa": 0.632, "hle": 0.052, "livecodebench": 0.561, "scicode": 0.331, "math_500": null, "aime": null, "aime_25": 0.443, "ifbench": 0.367, "lcr": 0.263, "terminalbench_hard": 0.288, "tau2": 0.769 }, "sub": true }, { "id": "deepseek-v3.2-exp", "name": "DeepSeek: DeepSeek V3.2 Exp", "short_name": "DeepSeek V3.2 Exp", "description": "DeepSeek-V3.2-Exp is an experimental large language model released by DeepSeek as an intermediate step between V3.1 and future architectures. It introduces DeepSeek Sparse Attention (DSA), a fine-grained sparse attention mechanism designed to improve training and inference efficiency in long-context scenarios while maintaining output quality. The model was trained under conditions aligned with V3.1-Terminus to enable direct comparison. Benchmarking shows performance roughly on par with V3.1 across reasoning, coding, and agentic tool-use tasks, with minor tradeoffs and gains depending on the domain. This release focuses on validating architectural optimizations for extended context lengths rather than advancing raw task accuracy, making it primarily a research-oriented model for exploring efficient transformer designs.", "context_length": 163840, "created": 1758548275, "owned_by": "deepseek-ai", "available": true, "hallucination_rate": 3.9, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.25 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.4 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": false, "function_call": true, "reasoning": false }, "benchmarks": { "artificial_analysis_intelligence_index": 28.4, "artificial_analysis_coding_index": 30, "artificial_analysis_math_index": 57.7, "mmlu_pro": 0.836, "gpqa": 0.738, "hle": 0.086, "livecodebench": 0.554, "scicode": 0.399, "math_500": null, "aime": null, "aime_25": 0.577, "ifbench": 0.431, "lcr": 0.43, "terminalbench_hard": 0.25, "tau2": 0.339 }, "sub": true }, { "id": "deepseek-v3.1-terminus", "name": "DeepSeek: DeepSeek V3.1 Terminus", "short_name": "DeepSeek V3.1 Terminus", "description": "DeepSeek-V3.1 Terminus is an update to [DeepSeek V3.1](/deepseek-ai/deepseek-v3.1) that maintains the model's original capabilities while addressing issues reported by users, including language consistency and agent capabilities, further optimizing the model's performance in coding and search agents. It is a large hybrid reasoning model (671B parameters, 37B active). It extends the DeepSeek-V3 base with a two-phase long-context training process, reaching up to 128K tokens, and uses FP8 microscaling for efficient inference.\nThe model improves tool use, code generation, and reasoning efficiency, achieving performance comparable to DeepSeek-R1 on difficult benchmarks while responding more quickly. It supports structured tool calling, code agents, and search agents, making it suitable for research, coding, and agentic workflows.", "context_length": 163840, "created": 1758548275, "owned_by": "deepseek-ai", "available": true, "hallucination_rate": 2, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.23 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.64 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": false, "function_call": true, "reasoning": false }, "benchmarks": { "artificial_analysis_intelligence_index": 28.5, "artificial_analysis_coding_index": 31.9, "artificial_analysis_math_index": 53.7, "mmlu_pro": 0.836, "gpqa": 0.751, "hle": 0.084, "livecodebench": 0.529, "scicode": 0.321, "math_500": null, "aime": null, "aime_25": 0.537, "ifbench": 0.412, "lcr": 0.433, "terminalbench_hard": 0.318, "tau2": 0.371 }, "sub": true }, { "id": "qwen3-235b-a22b-2507", "name": "Alibaba: Qwen3 235B A22B Instruct 2507", "short_name": "Qwen3 235B A22B Instruct 2507", "description": "A multilingual, instruction-tuned mixture-of-experts language model based on the Qwen3-235B architecture, with 22B active parameters per forward pass. It is optimized for general-purpose text generation, including instruction following, logical reasoning, math, code, and tool usage. The model supports a native 262K context length and does not implement 'thinking mode' ( blocks)", "context_length": 262000, "created": 1758210654, "owned_by": "z-ai", "available": true, "hallucination_rate": 45.7, "endpoints": [ "/v1/chat/completions" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.176 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.704 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": false, "function_call": true, "reasoning": false }, "benchmarks": { "artificial_analysis_intelligence_index": 25, "artificial_analysis_coding_index": 22.1, "artificial_analysis_math_index": 71.7, "mmlu_pro": 0.828, "gpqa": 0.753, "hle": 0.106, "livecodebench": 0.524, "scicode": 0.36, "math_500": 0.98, "aime": 0.717, "aime_25": 0.717, "ifbench": 0.461, "lcr": 0.312, "terminalbench_hard": 0.152, "tau2": 0.333 } }, { "id": "qwen3-235b-a22b-thinking-2507", "name": "Alibaba: Qwen3 235B A22B Thinking 2507", "short_name": "Qwen3 235B A22B Thinking 2507", "description": "A high-performance, open-weight Mixture-of-Experts (MoE) language model optimized for complex reasoning tasks. It activates 22B of its 235B parameters per forward pass and natively supports up to 262,144 tokens of context.", "context_length": 262000, "created": 1758210654, "owned_by": "z-ai", "available": true, "hallucination_rate": 20, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.176 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.704 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": false, "function_call": true, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": 25, "artificial_analysis_coding_index": 22.1, "artificial_analysis_math_index": 71.7, "mmlu_pro": 0.828, "gpqa": 0.753, "hle": 0.106, "livecodebench": 0.524, "scicode": 0.36, "math_500": 0.98, "aime": 0.717, "aime_25": 0.717, "ifbench": 0.461, "lcr": 0.312, "terminalbench_hard": 0.152, "tau2": 0.333 } }, { "id": "nemotron-nano-9b-v2:free", "name": "NVIDIA: Nemotron Nano 9B V2 (Free)", "short_name": "Nemotron Nano 9B V2 (Free)", "description": "NVIDIA-Nemotron-Nano-9B-v2 is a large language model (LLM) trained from scratch by NVIDIA, and designed as a unified model for both reasoning and non-reasoning tasks. It responds to user queries and tasks by first generating a reasoning trace and then concluding with a final response. The model's reasoning capabilities can be controlled via a system prompt. If the user prefers the model to provide its final answer without intermediate reasoning traces, it can be configured to do so.", "context_length": 128000, "created": 1757106807, "owned_by": "nvidia", "available": true, "hallucination_rate": 2.5, "endpoints": [ "/v1/chat/completions" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0 }, "output": { "unit": "1M tokens", "price_per_million_t": 0 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": false, "function_call": true, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": 13.2, "artificial_analysis_coding_index": 7.5, "artificial_analysis_math_index": 62.3, "mmlu_pro": 0.739, "gpqa": 0.557, "hle": 0.04, "livecodebench": 0.701, "scicode": 0.209, "math_500": null, "aime": null, "aime_25": 0.623, "ifbench": 0.271, "lcr": 0.227, "terminalbench_hard": 0.008, "tau2": 0.234 } }, { "id": "nemotron-nano-9b-v2", "name": "NVIDIA: Nemotron Nano 9B V2", "short_name": "Nemotron Nano 9B V2", "description": "NVIDIA-Nemotron-Nano-9B-v2 is a large language model (LLM) trained from scratch by NVIDIA, and designed as a unified model for both reasoning and non-reasoning tasks. It responds to user queries and tasks by first generating a reasoning trace and then concluding with a final response. The model's reasoning capabilities can be controlled via a system prompt. If the user prefers the model to provide its final answer without intermediate reasoning traces, it can be configured to do so.", "context_length": 128000, "created": 1757106807, "owned_by": "nvidia", "available": true, "hallucination_rate": 2.5, "endpoints": [ "/v1/chat/completions" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.02 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.04 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": false, "function_call": true, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": 13.2, "artificial_analysis_coding_index": 7.5, "artificial_analysis_math_index": 62.3, "mmlu_pro": 0.739, "gpqa": 0.557, "hle": 0.04, "livecodebench": 0.701, "scicode": 0.209, "math_500": null, "aime": null, "aime_25": 0.623, "ifbench": 0.271, "lcr": 0.227, "terminalbench_hard": 0.008, "tau2": 0.234 } }, { "id": "claude-opus-4-1", "name": "Anthropic: Claude Opus 4.1", "short_name": "Claude Opus 4.1", "description": "Claude Opus 4.1 is an updated version of Anthropic’s flagship model, offering improved performance in coding, reasoning, and agentic tasks. It achieves 74.5% on SWE-bench Verified and shows notable gains in multi-file code refactoring, debugging precision, and detail-oriented reasoning. The model supports extended thinking up to 64K tokens and is optimized for tasks involving research, data analysis, and tool-assisted reasoning.", "context_length": 200000, "created": 1754411591, "owned_by": "anthropic", "available": true, "hallucination_rate": 2.5, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 13.5 }, "output": { "unit": "1M tokens", "price_per_million_t": 67.5 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 1.5 }, "write": { "5m": { "unit": "1M tokens", "price_per_million_t": 18.74 }, "1h": { "unit": "1M tokens", "price_per_million_t": 29.99 } } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": true, "function_call": true, "reasoning": false }, "benchmarks": { "artificial_analysis_intelligence_index": 18, "artificial_analysis_coding_index": 19.5, "artificial_analysis_math_index": 62.7, "mmlu_pro": 0.696, "gpqa": 0.489, "hle": 0.031, "livecodebench": 0.279, "scicode": 0.233, "math_500": 0.641, "aime": 0.033, "aime_25": 0.627, "ifbench": 0.43, "lcr": 0.653, "terminalbench_hard": 0.409, "tau2": 0.863 }, "discount": { "percentage": 5, "note": "A 5% discount applies to input and output token pricing only." } }, { "id": "claude-opus-4", "name": "Anthropic: Claude Opus 4", "short_name": "Claude Opus 4", "description": "Claude Opus 4 is benchmarked as the world’s best coding model, at time of release, bringing sustained performance on complex, long-running tasks and agent workflows. It sets new benchmarks in software engineering, achieving leading results on SWE-bench (72.5%) and Terminal-bench (43.2%). Opus 4 supports extended, agentic workflows, handling thousands of task steps continuously for hours without degradation.", "context_length": 200000, "created": 1754411591, "owned_by": "anthropic", "available": true, "hallucination_rate": 2.5, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 13.5 }, "output": { "unit": "1M tokens", "price_per_million_t": 67.5 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 1.5 }, "write": { "5m": { "unit": "1M tokens", "price_per_million_t": 18.74 }, "1h": { "unit": "1M tokens", "price_per_million_t": 29.99 } } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": true, "function_call": true, "reasoning": false }, "benchmarks": { "artificial_analysis_intelligence_index": 18, "artificial_analysis_coding_index": 19.5, "artificial_analysis_math_index": 62.7, "mmlu_pro": 0.696, "gpqa": 0.489, "hle": 0.031, "livecodebench": 0.279, "scicode": 0.233, "math_500": 0.641, "aime": 0.033, "aime_25": 0.627, "ifbench": 0.43, "lcr": 0.653, "terminalbench_hard": 0.409, "tau2": 0.863 }, "discount": { "percentage": 5, "note": "A 5% discount applies to input and output token pricing only." } }, { "id": "aion-1.0-mini", "name": "AionLabs: Aion-1.0-Mini", "short_name": "Aion 1.0 Mini", "description": "Aion-1.0-Mini 32B parameter model is a distilled version of the DeepSeek-R1 model, designed for strong performance in reasoning domains such as mathematics, coding, and logic. It is a modified variant of a FuseAI model that outperforms R1-Distill-Qwen-32B and R1-Distill-Llama-70B, with benchmark results available on its [Hugging Face page](https://huggingface.co/FuseAI/FuseO1-DeepSeekR1-QwQ-SkyT1-32B-Preview), independently replicated for verification.", "context_length": 131072, "created": 1738697107, "owned_by": "aion-labs", "available": true, "hallucination_rate": 3.4, "endpoints": [ "/v1/chat/completions" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.2 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.28 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": false, "function_call": true, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": 21.7, "artificial_analysis_coding_index": 10.5, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": 0.657, "hle": 0.092, "livecodebench": null, "scicode": 0.269, "math_500": null, "aime": null, "aime_25": null, "ifbench": 0.577, "lcr": 0.36, "terminalbench_hard": 0.023, "tau2": 0.482 } }, { "id": "aion-rp-llama-3.1-8b", "name": "AionLabs: Aion-RP 1.0 8B", "short_name": "Aion-RP 1.0 8B", "description": "Aion-RP-Llama-3.1-8B ranks the highest in the character evaluation portion of the RPBench-Auto benchmark, a roleplaying-specific variant of Arena-Hard-Auto, where LLMs evaluate each other’s responses. It is a fine-tuned base model rather than an instruct model, designed to produce more natural and varied writing.", "context_length": 32768, "created": 1738696718, "owned_by": "aion-labs", "available": true, "hallucination_rate": 3.4, "endpoints": [ "/v1/chat/completions" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.2 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.28 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": false, "function_call": true, "reasoning": false } }, { "id": "deepseek-v3.1", "name": "DeepSeek: V3.1", "short_name": "DeepSeek V3.1", "description": "DeepSeek-V3.1 is a large hybrid reasoning model (671B parameters, 37B active) that supports both thinking and non-thinking modes via prompt templates. It extends the DeepSeek-V3 base with a two-phase long-context training process, reaching up to 128K tokens, and uses FP8 microscaling for efficient inference. The model improves tool use, code generation, and reasoning efficiency, achieving performance comparable to DeepSeek-R1 on difficult benchmarks while responding more quickly. It supports structured tool calling, code agents, and search agents, making it suitable for research, coding, and agentic workflows. It succeeds the DeepSeek V3-0324 model and performs well on a variety of tasks.", "context_length": 163840, "created": 1755779628, "owned_by": "deepseek-ai", "available": true, "hallucination_rate": 3.9, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.18 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.64 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": false, "function_call": true, "reasoning": false }, "benchmarks": { "artificial_analysis_intelligence_index": 28.1, "artificial_analysis_coding_index": 28.4, "artificial_analysis_math_index": 49.7, "mmlu_pro": 0.833, "gpqa": 0.735, "hle": 0.063, "livecodebench": 0.577, "scicode": 0.367, "aime_25": 0.497, "ifbench": 0.378, "lcr": 0.45, "terminalbench_hard": 0.242, "tau2": 0.348 }, "sub": true }, { "id": "gpt-oss-120b:free", "name": "OpenAI: GPT OSS 120B (Free)", "short_name": "GPT OSS 120B (Free)", "description": "gpt-oss-120b is an open-weight, 117B-parameter Mixture-of-Experts (MoE) language model from OpenAI designed for high-reasoning, agentic, and general-purpose production use cases. It activates 5.1B parameters per forward pass and is optimized to run on a single H100 GPU with native MXFP4 quantization. The model supports configurable reasoning depth, full chain-of-thought access, and native tool use, including function calling, browsing, and structured output generation.", "context_length": 131072, "created": 1754414231, "owned_by": "openai", "available": true, "hallucination_rate": 49.1, "endpoints": [ "/v1/chat/completions" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0 }, "output": { "unit": "1M tokens", "price_per_million_t": 0 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": false, "function_call": true, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": 24.5, "artificial_analysis_coding_index": 15.5, "artificial_analysis_math_index": 66.7, "mmlu_pro": 0.775, "gpqa": 0.672, "hle": 0.052, "livecodebench": 0.707, "scicode": 0.36, "math_500": null, "aime": null, "aime_25": 0.667, "ifbench": 0.583, "lcr": 0.437, "terminalbench_hard": 0.053, "tau2": 0.45 }, "discount": { "percentage": 5, "note": "A 5% discount applies to input and output token pricing only." } }, { "id": "gpt-oss-120b", "name": "OpenAI: GPT OSS 120B", "short_name": "GPT OSS 120B", "description": "gpt-oss-120b is an open-weight, 117B-parameter Mixture-of-Experts (MoE) language model from OpenAI designed for high-reasoning, agentic, and general-purpose production use cases. It activates 5.1B parameters per forward pass and is optimized to run on a single H100 GPU with native MXFP4 quantization. The model supports configurable reasoning depth, full chain-of-thought access, and native tool use, including function calling, browsing, and structured output generation.", "context_length": 131072, "created": 1754414231, "owned_by": "openai", "available": true, "hallucination_rate": 49.1, "endpoints": [ "/v1/chat/completions" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.04 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.3 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": false, "function_call": true, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": 24.5, "artificial_analysis_coding_index": 15.5, "artificial_analysis_math_index": 66.7, "mmlu_pro": 0.775, "gpqa": 0.672, "hle": 0.052, "livecodebench": 0.707, "scicode": 0.36, "math_500": null, "aime": null, "aime_25": 0.667, "ifbench": 0.583, "lcr": 0.437, "terminalbench_hard": 0.053, "tau2": 0.45 }, "discount": { "percentage": 5, "note": "A 5% discount applies to input and output token pricing only." } }, { "id": "gpt-oss-20b", "name": "OpenAI: GPT OSS 20B", "short_name": "GPT OSS 20B", "description": "gpt-oss-20b is an open-weight 21B parameter model released by OpenAI under the Apache 2.0 license. It uses a Mixture-of-Experts (MoE) architecture with 3.6B active parameters per forward pass, optimized for lower-latency inference and deployability on consumer or single-GPU hardware. The model is trained in OpenAI’s Harmony response format and supports reasoning level configuration, fine-tuning, and agentic capabilities including function calling, tool use, and structured outputs.", "context_length": 131072, "created": 1754414229, "owned_by": "openai", "available": true, "hallucination_rate": 49, "endpoints": [ "/v1/chat/completions" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.15 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.6 } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": false, "function_call": true, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": 24.5, "artificial_analysis_coding_index": 18.5, "artificial_analysis_math_index": 89.3, "mmlu_pro": 0.748, "gpqa": 0.688, "hle": 0.098, "livecodebench": 0.777, "scicode": 0.344, "math_500": null, "aime": null, "aime_25": 0.893, "ifbench": 0.651, "lcr": 0.307, "terminalbench_hard": 0.106, "tau2": 0.602 }, "discount": { "percentage": 5, "note": "A 5% discount applies to input and output token pricing only." } }, { "id": "glm-4.5-air", "name": "Z.AI: GLM 4.5 Air", "short_name": "GLM 4.5 Air", "description": "The GLM-4.5 series models are foundation models designed for intelligent agents. GLM-4.5 has 355 billion total parameters with 32 billion active parameters, while GLM-4.5-Air adopts a more compact design with 106 billion total parameters and 12 billion active parameters. GLM-4.5 models unify reasoning, coding, and intelligent agent capabilities to meet the complex demands of intelligent agent applications.", "context_length": 131072, "created": 1753471258, "owned_by": "z-ai", "available": true, "hallucination_rate": 1.3, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.1 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.1 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0.08 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": false, "function_call": true, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": 23.2, "artificial_analysis_coding_index": 23.8, "artificial_analysis_math_index": 80.7, "mmlu_pro": 0.815, "gpqa": 0.733, "hle": 0.068, "livecodebench": 0.684, "scicode": 0.306, "math_500": 0.965, "aime": 0.673, "aime_25": 0.807, "ifbench": 0.376, "lcr": 0.437, "terminalbench_hard": 0.205, "tau2": 0.465 }, "sub": true }, { "id": "gpt-5", "name": "OpenAI: GPT-5", "short_name": "GPT-5", "description": "GPT-5 is OpenAI’s most advanced model, offering major improvements in reasoning, code quality, and user experience. It is optimized for complex tasks that require step-by-step reasoning, instruction following, and accuracy in high-stakes use cases.", "context_length": 400000, "created": 1753471258, "owned_by": "openai", "available": true, "hallucination_rate": 1.3, "endpoints": [ "/v1/chat/completions" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.625 }, "output": { "unit": "1M tokens", "price_per_million_t": 5 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0.13 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": true, "function_call": true, "reasoning": true, "supported_reasoning_efforts": [ "minimal", "low", "medium", "high" ] }, "benchmarks": { "artificial_analysis_intelligence_index": 23.9, "artificial_analysis_coding_index": 25.1, "artificial_analysis_math_index": 31.7, "mmlu_pro": 0.806, "gpqa": 0.673, "hle": 0.054, "livecodebench": 0.558, "scicode": 0.388, "math_500": 0.861, "aime": 0.367, "aime_25": 0.317, "ifbench": 0.456, "lcr": 0.25, "terminalbench_hard": 0.182, "tau2": 0.67 }, "discount": { "percentage": 5, "note": "A 5% discount applies to input and output token pricing only." } }, { "id": "gpt-5-mini", "name": "OpenAI: GPT-5 Mini", "short_name": "GPT-5 Mini", "description": "GPT-5 Mini is a compact version of GPT-5, designed to handle lighter-weight reasoning tasks. It provides the same instruction-following and safety-tuning benefits as GPT-5, but with reduced latency and cost. GPT-5 Mini is the successor to OpenAI's o4-mini model.", "context_length": 400000, "created": 1753471258, "owned_by": "openai", "available": true, "hallucination_rate": 1.3, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.125 }, "output": { "unit": "1M tokens", "price_per_million_t": 1 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0.03 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": true, "function_call": true, "reasoning": true, "supported_reasoning_efforts": [ "minimal", "low", "medium", "high" ] }, "benchmarks": { "artificial_analysis_intelligence_index": 20.7, "artificial_analysis_coding_index": 21.9, "artificial_analysis_math_index": 46.7, "mmlu_pro": 0.775, "gpqa": 0.687, "hle": 0.05, "livecodebench": 0.545, "scicode": 0.369, "math_500": null, "aime": null, "aime_25": 0.467, "ifbench": 0.456, "lcr": 0.357, "terminalbench_hard": 0.144, "tau2": 0.319 }, "discount": { "percentage": 5, "note": "A 5% discount applies to input and output token pricing only." } }, { "id": "trinity-large-thinking", "name": "Arcee AI: Trinity Large Thinking", "short_name": "Trinity Large Thinking", "description": "Trinity Large Thinking is a powerful open source reasoning model from the team at Arcee AI. It shows strong performance in PinchBench, agentic workloads, and reasoning tasks", "context_length": 131000, "created": 1775001600, "owned_by": "arceeai", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.25 }, "output": { "unit": "1M tokens", "price_per_million_t": 1 } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": false, "function_call": true, "reasoning": false }, "benchmarks": { "artificial_analysis_intelligence_index": 32.5, "artificial_analysis_coding_index": 24.5, "artificial_analysis_math_index": 82.3, "mmlu_pro": 0.824, "gpqa": 0.776, "hle": 0.12, "livecodebench": 0.535, "scicode": 0.387, "math_500": null, "aime": null, "aime_25": 0.823, "ifbench": 0.538, "lcr": 0.577, "terminalbench_hard": 0.174, "tau2": 0.836 }, "sub": true }, { "id": "gpt-5-nano", "name": "OpenAI: GPT-5-Nano", "short_name": "GPT-5-Nano", "description": "GPT-5-Nano is the smallest and fastest variant in the GPT-5 system, optimized for developer tools, rapid interactions, and ultra-low latency environments.", "context_length": 400000, "created": 1753471258, "owned_by": "openai", "available": true, "hallucination_rate": 1.3, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.05 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.2 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0.005 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": true, "function_call": true, "reasoning": true, "supported_reasoning_efforts": [ "minimal", "low", "medium", "high" ] }, "benchmarks": { "artificial_analysis_intelligence_index": 13.8, "artificial_analysis_coding_index": 14.2, "artificial_analysis_math_index": 27.3, "mmlu_pro": 0.556, "gpqa": 0.428, "hle": 0.041, "livecodebench": 0.47, "scicode": 0.291, "math_500": null, "aime": null, "aime_25": 0.273, "ifbench": 0.325, "lcr": 0.2, "terminalbench_hard": 0.068, "tau2": 0.257 }, "discount": { "percentage": 5, "note": "A 5% discount applies to input and output token pricing only." } }, { "id": "llama-4-maverick", "name": "Meta: Llama 4 Maverick", "short_name": "Llama 4 Maverick", "description": "Llama 4 Maverick 17B Instruct (128E) is a high-capacity multimodal language model from Meta, built on a mixture-of-experts (MoE) architecture with 128 experts and 17 billion active parameters per forward pass (400B total). It supports multilingual text and image input, and produces multilingual text and code output across 12 supported languages. Optimized for vision-language tasks, Maverick is instruction-tuned for assistant-like behavior, image reasoning, and general-purpose multimodal interaction.\n\nMaverick features early fusion for native multimodality and a 1 million token context window. It was trained on a curated mixture of public, licensed, and Meta-platform data, covering ~22 trillion tokens, with a knowledge cutoff in August 2024. Released on April 5, 2025 under the Llama 4 Community License, Maverick is suited for research and commercial applications requiring advanced multimodal understanding and high model throughput.", "context_length": 1048576, "created": 1743881822, "owned_by": "meta", "available": true, "hallucination_rate": 5.56, "endpoints": [ "/v1/chat/completions" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.11 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.44 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0.02 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": true, "function_call": true, "reasoning": false }, "benchmarks": { "artificial_analysis_intelligence_index": 18.4, "artificial_analysis_coding_index": 15.6, "artificial_analysis_math_index": 19.3, "mmlu_pro": 0.809, "gpqa": 0.671, "hle": 0.048, "livecodebench": 0.397, "scicode": 0.331, "math_500": 0.889, "aime": 0.39, "aime_25": 0.193, "ifbench": 0.43, "lcr": 0.46, "terminalbench_hard": 0.068, "tau2": 0.178 } }, { "id": "llama-4-scout", "name": "Meta: Llama 4 Scout", "short_name": "Llama 4 Scout", "description": "Llama 4 Scout 17B Instruct (16E) is a mixture-of-experts (MoE) language model developed by Meta, activating 17 billion parameters out of a total of 109B. It supports native multimodal input (text and image) and multilingual output (text and code) across 12 supported languages. Designed for assistant-style interaction and visual reasoning, Scout uses 16 experts per forward pass and features a context length of 10 million tokens, with a training corpus of ~40 trillion tokens.\n\nBuilt for high efficiency and local or commercial deployment, Llama 4 Scout incorporates early fusion for seamless modality integration. It is instruction-tuned for use in multilingual chat, captioning, and image understanding tasks. Released under the Llama 4 Community License, it was last trained on data up to August 2024 and launched publicly on April 5, 2025.", "context_length": 1048576, "created": 1743881519, "owned_by": "meta", "available": true, "hallucination_rate": 1.62, "endpoints": [ "/v1/chat/completions" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.1 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.225 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": true, "function_call": true, "reasoning": false }, "benchmarks": { "artificial_analysis_intelligence_index": 13.5, "artificial_analysis_coding_index": 6.7, "artificial_analysis_math_index": 14, "mmlu_pro": 0.752, "gpqa": 0.587, "hle": 0.043, "livecodebench": 0.299, "scicode": 0.17, "math_500": 0.844, "aime": 0.283, "aime_25": 0.14, "ifbench": 0.395, "lcr": 0.258, "terminalbench_hard": 0.015, "tau2": 0.155 } }, { "id": "gpt-4.1", "name": "OpenAI: GPT-4.1", "short_name": "GPT-4.1", "description": "GPT-4.1 is a flagship large language model optimized for advanced instruction following, real-world software engineering, and long-context reasoning. It supports a 1 million token context window and outperforms GPT-4o and GPT-4.5 across coding (54.6% SWE-bench Verified), instruction compliance (87.4% IFEval), and multimodal understanding benchmarks. It is tuned for precise code diffs, agent reliability, and high recall in large document contexts, making it ideal for agents, IDE tooling, and enterprise knowledge retrieval.", "context_length": 1047576, "created": 1744651385, "owned_by": "openai", "available": true, "hallucination_rate": 2.8, "endpoints": [ "/v1/chat/completions" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 1 }, "output": { "unit": "1M tokens", "price_per_million_t": 4 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0.5 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": true, "function_call": true, "reasoning": false }, "benchmarks": { "artificial_analysis_intelligence_index": 26.3, "artificial_analysis_coding_index": 21.8, "artificial_analysis_math_index": 34.7, "mmlu_pro": 0.806, "gpqa": 0.666, "hle": 0.046, "livecodebench": 0.457, "scicode": 0.381, "math_500": 0.913, "aime": 0.437, "aime_25": 0.347, "ifbench": 0.43, "lcr": 0.61, "terminalbench_hard": 0.136, "tau2": 0.471 }, "discount": { "percentage": 5, "note": "A 5% discount applies to input and output token pricing only." } }, { "id": "gpt-4.1-nano", "name": "OpenAI: GPT-4.1 Nano", "short_name": "GPT-4.1 Nano", "description": "For tasks that demand low latency, GPT‑4.1 nano is the fastest and cheapest model in the GPT-4.1 series. It delivers exceptional performance at a small size with its 1 million token context window, and scores 80.1% on MMLU, 50.3% on GPQA, and 9.8% on Aider polyglot coding – even higher than GPT‑4o mini. It’s ideal for tasks like classification or autocompletion.", "context_length": 1047576, "created": 1744651369, "owned_by": "openai", "available": true, "hallucination_rate": 2, "endpoints": [ "/v1/chat/completions" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.05 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.2 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0.03 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": true, "function_call": true, "reasoning": false }, "benchmarks": { "artificial_analysis_intelligence_index": 13, "artificial_analysis_coding_index": 11.2, "artificial_analysis_math_index": 24, "mmlu_pro": 0.657, "gpqa": 0.512, "hle": 0.039, "livecodebench": 0.326, "scicode": 0.259, "math_500": 0.848, "aime": 0.237, "aime_25": 0.24, "ifbench": 0.32, "lcr": 0.17, "terminalbench_hard": 0.038, "tau2": 0.173 }, "discount": { "percentage": 5, "note": "A 5% discount applies to input and output token pricing only." } }, { "id": "o3-mini", "name": "OpenAI: o3 Mini", "short_name": "o3 Mini", "description": "A smaller variant of O3, expected to offer enhanced multimodal capabilities, improved reasoning, and more efficient resource utilization compared to previous models while maintaining strong performance on core tasks.", "context_length": 200000, "created": 1738351721, "owned_by": "openai", "available": true, "hallucination_rate": 14.8, "endpoints": [ "/v1/chat/completions" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 1.09 }, "output": { "unit": "1M tokens", "price_per_million_t": 4.4 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0.54 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": false, "function_call": true, "reasoning": true, "supported_reasoning_efforts": [ "low", "medium", "high" ] }, "benchmarks": { "artificial_analysis_intelligence_index": 25.9, "artificial_analysis_coding_index": 17.9, "artificial_analysis_math_index": null, "mmlu_pro": 0.791, "gpqa": 0.748, "hle": 0.087, "livecodebench": 0.717, "scicode": 0.399, "math_500": 0.973, "aime": 0.77, "aime_25": null, "ifbench": null, "lcr": null, "terminalbench_hard": 0.068, "tau2": 0.287 }, "discount": { "percentage": 5, "note": "A 5% discount applies to input and output token pricing only." } }, { "id": "deepseek-r1-0528", "name": "DeepSeek: R1 0528", "short_name": "DeepSeek R1 0528", "description": "DeepSeek-R1 is the first-generation reasoning model built atop DeepSeek-V3 (671B total parameters, 37B activated per token). It incorporates large-scale reinforcement learning (RL) to enhance its chain-of-thought and reasoning capabilities, delivering strong performance in math, code, and multi-step reasoning tasks.", "context_length": 163840, "created": 1737381095, "owned_by": "deepseek-ai", "available": true, "hallucination_rate": 14.3, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.37 }, "output": { "unit": "1M tokens", "price_per_million_t": 1.56 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": false, "function_call": true, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": 27.1, "artificial_analysis_coding_index": 24, "artificial_analysis_math_index": 76, "mmlu_pro": 0.849, "gpqa": 0.813, "hle": 0.149, "livecodebench": 0.77, "scicode": 0.403, "math_500": 0.983, "aime": 0.893, "aime_25": 0.76, "ifbench": 0.396, "lcr": 0.547, "terminalbench_hard": 0.159, "tau2": 0.365 } }, { "id": "deepseek-v3-0324", "name": "DeepSeek: DeepSeek V3 0324", "short_name": "DeepSeek V3 0324", "description": "A powerful Mixture-of-Experts (MoE) language model with 671B total parameters (37B activated per token). Features Multi-head Latent Attention (MLA), auxiliary-loss-free load balancing, and multi-token prediction training. Pre-trained on 14.8T tokens with strong performance in reasoning, math, and code tasks.", "context_length": 163840, "created": 1742824755, "owned_by": "deepseek-ai", "available": true, "hallucination_rate": 3.9, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.23 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.64 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": false, "function_call": true, "reasoning": false }, "benchmarks": { "artificial_analysis_intelligence_index": 22.3, "artificial_analysis_coding_index": 22, "artificial_analysis_math_index": 41, "mmlu_pro": 0.819, "gpqa": 0.655, "hle": 0.052, "livecodebench": 0.405, "scicode": 0.358, "math_500": 0.942, "aime": 0.52, "aime_25": 0.41, "ifbench": 0.41, "lcr": 0.41, "terminalbench_hard": 0.152, "tau2": 0.471 }, "sub": true }, { "id": "deepseek-r1-distill-qwen-32b", "name": "DeepSeek: R1 Distill Qwen 32B", "short_name": "R1 Distill Qwen 32B", "description": "DeepSeek R1 Distill Qwen 32B is a sophisticated language model designed for users who need advanced reasoning abilities. This model excels at complex problem-solving, demonstrating remarkable proficiency in tasks requiring logical deduction and critical thinking. It utilizes a distillation architecture, meaning it has learned to perform like a much larger model but in a more efficient package. DeepSeek R1 Distill Qwen 32B achieves impressive results on reasoning benchmarks, even surpassing larger models in certain areas. A standout feature is its ability to perform at a high level while being more accessible for deployment.", "context_length": 131072, "created": 1738194830, "owned_by": "deepseek-ai", "available": true, "hallucination_rate": 14.3, "endpoints": [ "/v1/chat/completions" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.3 }, "output": { "unit": "1M tokens", "price_per_million_t": 2.928 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": false, "function_call": true, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": 17.2, "artificial_analysis_coding_index": null, "artificial_analysis_math_index": 63, "mmlu_pro": 0.739, "gpqa": 0.615, "hle": 0.055, "livecodebench": 0.27, "scicode": 0.376, "math_500": 0.941, "aime": 0.687, "aime_25": 0.63, "ifbench": 0.229, "lcr": 0.097, "terminalbench_hard": null, "tau2": null } }, { "id": "gpt-4o", "name": "OpenAI: GPT-4o", "short_name": "GPT-4o", "description": "GPT-4o (\"o\" for \"omni\") is OpenAI's latest AI model, supporting both text and image inputs with text outputs. It maintains the intelligence level of [GPT-4 Turbo](/models/openai/gpt-4-turbo) while being twice as fast and 50% more cost-effective. GPT-4o also offers improved performance in processing non-English languages and enhanced visual capabilities.\n\nFor benchmarking against other models, it was briefly called [\"im-also-a-good-gpt2-chatbot\"](https://twitter.com/LiamFedus/status/1790064963966370209)\n\n#multimodal", "context_length": 128000, "created": 1715558400, "owned_by": "openai", "available": true, "hallucination_rate": 1.5, "endpoints": [ "/v1/chat/completions" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 2.5 }, "output": { "unit": "1M tokens", "price_per_million_t": 10 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 1.25 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": true, "function_call": true, "reasoning": false }, "benchmarks": { "artificial_analysis_intelligence_index": 18.6, "artificial_analysis_coding_index": 16.6, "artificial_analysis_math_index": 6, "mmlu_pro": 0.748, "gpqa": 0.521, "hle": 0.029, "livecodebench": 0.317, "scicode": 0.331, "math_500": 0.795, "aime": 0.117, "aime_25": 0.06, "ifbench": 0.36, "lcr": 0.35, "terminalbench_hard": 0.083, "tau2": 0.289 }, "discount": { "percentage": 5, "note": "A 5% discount applies to input and output token pricing only." } }, { "id": "gpt-4o-mini", "name": "OpenAI: GPT-4o-mini", "short_name": "GPT-4o-mini", "description": "GPT-4o mini is OpenAI's newest model (released July 18, 2024), following GPT-4 Omni. It supports both text and image inputs, producing text outputs. As OpenAI's most advanced small model, it offers state-of-the-art intelligence at a fraction of the cost—over 60% cheaper than GPT-3.5 Turbo—making it vastly more affordable than other recent frontier models", "context_length": 128000, "created": 1721260800, "owned_by": "openai", "available": true, "hallucination_rate": 1.7, "endpoints": [ "/v1/chat/completions" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.4 }, "output": { "unit": "1M tokens", "price_per_million_t": 1.6 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0.1 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": true, "function_call": true, "reasoning": false }, "benchmarks": { "artificial_analysis_intelligence_index": 12.6, "artificial_analysis_coding_index": null, "artificial_analysis_math_index": 14.7, "mmlu_pro": 0.648, "gpqa": 0.426, "hle": 0.04, "livecodebench": 0.234, "scicode": 0.229, "math_500": 0.789, "aime": 0.117, "aime_25": 0.147, "ifbench": 0.31, "lcr": null, "terminalbench_hard": null, "tau2": null }, "discount": { "percentage": 5, "note": "A 5% discount applies to input and output token pricing only." } }, { "id": "llama-3.2-3b-instruct:free", "name": "Meta: Llama 3.2 3B Instruct (Free)", "short_name": "Llama 3.2 3B Instruct (Free)", "description": "Llama 3.2 3B is a 3-billion-parameter multilingual large language model, optimized for advanced natural language processing tasks like dialogue generation, reasoning, and summarization. Designed with the latest transformer architecture, it supports eight languages, including English, Spanish, and Hindi, and is adaptable for additional languages.\n\nTrained on 9 trillion tokens, the Llama 3.2 3B model excels in instruction-following, complex reasoning, and tool use. Its balanced performance makes it ideal for applications needing accuracy and efficiency in text generation across multilingual settings.\n\nClick here for the [original model card](https://github.com/meta-llama/llama-models/blob/main/models/llama3_2/MODEL_CARD.md).\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://www.llama.com/llama3/use-policy/).", "context_length": 16384, "created": 1727222400, "owned_by": "meta", "available": true, "hallucination_rate": 48.37, "endpoints": [ "/v1/chat/completions" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0 }, "output": { "unit": "1M tokens", "price_per_million_t": 0 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": false, "function_call": false, "reasoning": false }, "benchmarks": { "artificial_analysis_intelligence_index": 9.7, "artificial_analysis_coding_index": 10.7, "artificial_analysis_math_index": 3.3, "mmlu_pro": 0.347, "gpqa": 0.255, "hle": 0.052, "livecodebench": 0.083, "scicode": 0.052, "math_500": 0.489, "aime": 0.067, "aime_25": 0.033, "ifbench": 0.262, "lcr": 0.02, "terminalbench_hard": 0.03, "tau2": 0.211 } }, { "id": "llama-3.2-3b-instruct", "name": "Meta: Llama 3.2 3B Instruct", "short_name": "Llama 3.2 3B Instruct", "description": "Llama 3.2 3B is a 3-billion-parameter multilingual large language model, optimized for advanced natural language processing tasks like dialogue generation, reasoning, and summarization. Designed with the latest transformer architecture, it supports eight languages, including English, Spanish, and Hindi, and is adaptable for additional languages.\n\nTrained on 9 trillion tokens, the Llama 3.2 3B model excels in instruction-following, complex reasoning, and tool use. Its balanced performance makes it ideal for applications needing accuracy and efficiency in text generation across multilingual settings.\n\nClick here for the [original model card](https://github.com/meta-llama/llama-models/blob/main/models/llama3_2/MODEL_CARD.md).\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://www.llama.com/llama3/use-policy/).", "context_length": 16384, "created": 1727222400, "owned_by": "meta", "available": true, "hallucination_rate": 48.37, "endpoints": [ "/v1/chat/completions" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.02 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.05 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": false, "function_call": false, "reasoning": false }, "benchmarks": { "artificial_analysis_intelligence_index": 9.7, "artificial_analysis_coding_index": 10.7, "artificial_analysis_math_index": 3.3, "mmlu_pro": 0.347, "gpqa": 0.255, "hle": 0.052, "livecodebench": 0.083, "scicode": 0.052, "math_500": 0.489, "aime": 0.067, "aime_25": 0.033, "ifbench": 0.262, "lcr": 0.02, "terminalbench_hard": 0.03, "tau2": 0.211 } }, { "id": "llama-3.2-1b-instruct:free", "name": "Meta: Llama 3.2 1B Instruct (Free)", "short_name": "Llama 3.2 1B Instruct (Free)", "description": "Llama 3.2 1B is a 1-billion-parameter language model focused on efficiently performing natural language tasks, such as summarization, dialogue, and multilingual text analysis. Its smaller size allows it to operate efficiently in low-resource environments while maintaining strong task performance.\n\nSupporting eight core languages and fine-tunable for more, Llama 1.3B is ideal for businesses or developers seeking lightweight yet powerful AI solutions that can operate in diverse multilingual settings without the high computational demand of larger models.\n\nClick here for the [original model card](https://github.com/meta-llama/llama-models/blob/main/models/llama3_2/MODEL_CARD.md).\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://www.llama.com/llama3/use-policy/).", "context_length": 16384, "created": 1727222400, "owned_by": "meta", "available": true, "hallucination_rate": 48.37, "endpoints": [ "/v1/chat/completions" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0 }, "output": { "unit": "1M tokens", "price_per_million_t": 0 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": false, "function_call": false, "reasoning": false }, "benchmarks": { "artificial_analysis_intelligence_index": 11.8, "artificial_analysis_coding_index": 4.9, "artificial_analysis_math_index": 4.3, "mmlu_pro": 0.476, "gpqa": 0.259, "hle": 0.051, "livecodebench": 0.116, "scicode": 0.132, "math_500": 0.519, "aime": 0.077, "aime_25": 0.043, "ifbench": 0.286, "lcr": 0.157, "terminalbench_hard": 0.008, "tau2": 0.164 } }, { "id": "llama-3.2-1b-instruct", "name": "Meta: Llama 3.2 1B Instruct", "short_name": "Llama 3.2 1B Instruct", "description": "Llama 3.2 1B is a 1-billion-parameter language model focused on efficiently performing natural language tasks, such as summarization, dialogue, and multilingual text analysis. Its smaller size allows it to operate efficiently in low-resource environments while maintaining strong task performance.\n\nSupporting eight core languages and fine-tunable for more, Llama 1.3B is ideal for businesses or developers seeking lightweight yet powerful AI solutions that can operate in diverse multilingual settings without the high computational demand of larger models.\n\nClick here for the [original model card](https://github.com/meta-llama/llama-models/blob/main/models/llama3_2/MODEL_CARD.md).\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://www.llama.com/llama3/use-policy/).", "context_length": 16384, "created": 1727222400, "owned_by": "meta", "available": true, "hallucination_rate": 48.37, "endpoints": [ "/v1/chat/completions" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.15 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.07 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": false, "function_call": false, "reasoning": false }, "benchmarks": { "artificial_analysis_intelligence_index": 11.8, "artificial_analysis_coding_index": 4.9, "artificial_analysis_math_index": 4.3, "mmlu_pro": 0.476, "gpqa": 0.259, "hle": 0.051, "livecodebench": 0.116, "scicode": 0.132, "math_500": 0.519, "aime": 0.077, "aime_25": 0.043, "ifbench": 0.286, "lcr": 0.157, "terminalbench_hard": 0.008, "tau2": 0.164 } }, { "id": "llama-3.1-8b-instruct:free", "name": "Meta: Llama 3.1 8B Instruct (Free)", "short_name": "Llama 3.1 8B Instruct (Free)", "description": "Meta's latest class of model (Llama 3.1) launched with a variety of sizes & flavors. This 8B instruct-tuned version is fast and efficient. It has demonstrated strong performance compared to leading closed-source models in human evaluations.", "context_length": 16384, "created": 1721692800, "owned_by": "meta", "available": true, "hallucination_rate": 48.37, "endpoints": [ "/v1/chat/completions" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0 }, "output": { "unit": "1M tokens", "price_per_million_t": 0 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": false, "function_call": false, "reasoning": false }, "benchmarks": { "artificial_analysis_intelligence_index": 11.8, "artificial_analysis_coding_index": 4.9, "artificial_analysis_math_index": 4.3, "mmlu_pro": 0.476, "gpqa": 0.259, "hle": 0.051, "livecodebench": 0.116, "scicode": 0.132, "math_500": 0.519, "aime": 0.077, "aime_25": 0.043, "ifbench": 0.286, "lcr": 0.157, "terminalbench_hard": 0.008, "tau2": 0.164 } }, { "id": "llama-3.1-8b-instruct", "name": "Meta: Llama 3.1 8B Instruct", "short_name": "Llama 3.1 8B Instruct", "description": "Meta's latest class of model (Llama 3.1) launched with a variety of sizes & flavors. This 8B instruct-tuned version is fast and efficient. It has demonstrated strong performance compared to leading closed-source models in human evaluations.", "context_length": 16384, "created": 1721692800, "owned_by": "meta", "available": true, "hallucination_rate": 48.37, "endpoints": [ "/v1/chat/completions" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.09 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.09 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": false, "function_call": false, "reasoning": false }, "benchmarks": { "artificial_analysis_intelligence_index": 11.8, "artificial_analysis_coding_index": 4.9, "artificial_analysis_math_index": 4.3, "mmlu_pro": 0.476, "gpqa": 0.259, "hle": 0.051, "livecodebench": 0.116, "scicode": 0.132, "math_500": 0.519, "aime": 0.077, "aime_25": 0.043, "ifbench": 0.286, "lcr": 0.157, "terminalbench_hard": 0.008, "tau2": 0.164 } }, { "id": "llama-3.1-405b-instruct", "name": "Meta: Llama 3.1 405B", "short_name": "Llama 3.1 405B", "description": "Meta's latest class of model (Llama 3.1) launched with a variety of sizes & flavors. This is the base 405B pre-trained version. It has demonstrated strong performance compared to leading closed-source models in human evaluations.", "context_length": 65536, "created": 1721692800, "owned_by": "meta", "available": true, "hallucination_rate": 26.84, "endpoints": [ "/v1/chat/completions" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 1.5 }, "output": { "unit": "1M tokens", "price_per_million_t": 1.5 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": false, "function_call": false, "reasoning": false }, "benchmarks": { "artificial_analysis_intelligence_index": 17.4, "artificial_analysis_coding_index": 14.5, "artificial_analysis_math_index": 3, "mmlu_pro": 0.732, "gpqa": 0.515, "hle": 0.042, "livecodebench": 0.305, "scicode": 0.299, "math_500": 0.703, "aime": 0.213, "aime_25": 0.03, "ifbench": 0.39, "lcr": 0.243, "terminalbench_hard": 0.068, "tau2": 0.19 } }, { "id": "llama-3.3-70b-instruct", "name": "Meta: Llama 3.3 70B Instruct", "short_name": "Llama 3.3 70B Instruct", "description": "The Meta Llama 3.3 is a 70B-parameter multilingual LLM, pretrained and instruction-tuned for text-only input and output. Optimized for multilingual dialogue, it outperforms many open and closed models on key benchmarks.", "context_length": 131072, "created": 1733506137, "owned_by": "meta", "available": true, "hallucination_rate": 50.19, "endpoints": [ "/v1/chat/completions" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.13 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.39 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": false, "function_call": true, "reasoning": false }, "benchmarks": { "artificial_analysis_intelligence_index": 14.5, "artificial_analysis_coding_index": 10.7, "artificial_analysis_math_index": 7.7, "mmlu_pro": 0.713, "gpqa": 0.498, "hle": 0.04, "livecodebench": 0.288, "scicode": 0.26, "math_500": 0.773, "aime": 0.3, "aime_25": 0.077, "ifbench": 0.471, "lcr": 0.15, "terminalbench_hard": 0.03, "tau2": 0.266 } }, { "id": "llama-3.3-70b-instruct:free", "name": "Meta: Llama 3.3 70B Instruct (Free)", "short_name": "Llama 3.3 70B Instruct (Free)", "description": "The Meta Llama 3.3 is a 70B-parameter multilingual LLM, pretrained and instruction-tuned for text-only input and output. Optimized for multilingual dialogue, it outperforms many open and closed models on key benchmarks.", "context_length": 131072, "created": 1733506137, "owned_by": "meta", "available": true, "hallucination_rate": 50.19, "endpoints": [ "/v1/chat/completions" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0 }, "output": { "unit": "1M tokens", "price_per_million_t": 0 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": false, "function_call": false, "reasoning": false }, "benchmarks": { "artificial_analysis_intelligence_index": 14.5, "artificial_analysis_coding_index": 10.7, "artificial_analysis_math_index": 7.7, "mmlu_pro": 0.713, "gpqa": 0.498, "hle": 0.04, "livecodebench": 0.288, "scicode": 0.26, "math_500": 0.773, "aime": 0.3, "aime_25": 0.077, "ifbench": 0.471, "lcr": 0.15, "terminalbench_hard": 0.03, "tau2": 0.266 } }, { "id": "qwq-32b", "name": "Qwen: QwQ 32B", "short_name": "QwQ 32B", "description": "QwQ-32B is a 32-billion-parameter AI model from Alibaba, built for advanced reasoning, coding, and problem-solving, featuring strong performance due to reinforcement learning and a very large 131,072-token context window", "context_length": 131072, "created": 1741123200, "owned_by": "qwen", "available": true, "hallucination_rate": 15, "endpoints": [ "/v1/chat/completions" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.5 }, "output": { "unit": "1M tokens", "price_per_million_t": 1 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": false, "function_call": true, "reasoning": true }, "benchmarks": { "artificial_analysis_intelligence_index": 19.7, "artificial_analysis_coding_index": null, "artificial_analysis_math_index": 29, "mmlu_pro": 0.764, "gpqa": 0.593, "hle": 0.082, "livecodebench": 0.631, "scicode": 0.358, "math_500": 0.957, "aime": 0.78, "aime_25": 0.29, "ifbench": 0.388, "lcr": 0.25, "terminalbench_hard": null, "tau2": null } }, { "id": "gemini-2.5-pro", "name": "Google: Gemini 2.5 Pro", "short_name": "Gemini 2.5 Pro", "description": "Gemini 2.5 Pro leads on common benchmarks with enhanced reasoning, multimodal capabilities, and a 1M token context window.", "context_length": 1048576, "created": 1750169544, "owned_by": "google", "available": true, "hallucination_rate": 3.2, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 2.25 }, "output": { "unit": "1M tokens", "price_per_million_t": 9 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0.25 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": true, "function_call": true, "reasoning": false }, "benchmarks": { "artificial_analysis_intelligence_index": 34.6, "artificial_analysis_coding_index": 31.9, "artificial_analysis_math_index": 87.7, "mmlu_pro": 0.862, "gpqa": 0.844, "hle": 0.211, "livecodebench": 0.801, "scicode": 0.428, "math_500": 0.967, "aime": 0.887, "aime_25": 0.877, "ifbench": 0.487, "lcr": 0.66, "terminalbench_hard": 0.265, "tau2": 0.541 }, "discount": { "percentage": 5, "note": "A 5% discount applies to input and output token pricing only." } }, { "id": "gemini-2.5-flash", "name": "Google: Gemini 2.5 Flash", "short_name": "Gemini 2.5 Flash", "description": "A thinking model designed for a balance between price and performance. It builds upon Gemini 2.0 Flash with upgraded reasoning, hybrid thinking control, multimodal capabilities, and a 1M token input context window.", "context_length": 1048576, "created": 1750172488, "owned_by": "google", "available": true, "hallucination_rate": 1.3, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.27 }, "output": { "unit": "1M tokens", "price_per_million_t": 2.25 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0.03 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": true, "function_call": true, "reasoning": false }, "benchmarks": { "artificial_analysis_intelligence_index": 27, "artificial_analysis_coding_index": 22.2, "artificial_analysis_math_index": 73.3, "mmlu_pro": 0.832, "gpqa": 0.79, "hle": 0.111, "livecodebench": 0.695, "scicode": 0.394, "math_500": 0.981, "aime": 0.823, "aime_25": 0.733, "ifbench": 0.503, "lcr": 0.617, "terminalbench_hard": 0.136, "tau2": 0.316 }, "discount": { "percentage": 5, "note": "A 5% discount applies to input and output token pricing only." } }, { "id": "mistral-nemo-instruct", "name": "Mistral: Mistral Nemo Instruct", "short_name": "Mistral Nemo Instruct", "description": "A 12B parameter model built by Mistral in collaboration with NVIDIA.\n\nThe model is multilingual, supporting English, French, German, Spanish, Italian, Portuguese, Chinese, Japanese, Korean, Arabic, and Hindi.\nIt supports function calling and is released under the Apache 2.0 license.", "context_length": 16384, "created": 1721347200, "owned_by": "mistral", "available": true, "hallucination_rate": 11.2, "endpoints": [ "/v1/chat/completions" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.02 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.04 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": false, "function_call": false, "reasoning": false }, "benchmarks": { "artificial_analysis_intelligence_index": 7.4, "artificial_analysis_coding_index": null, "artificial_analysis_math_index": null, "mmlu_pro": 0.245, "gpqa": 0.177, "hle": 0.043, "livecodebench": 0.046, "scicode": 0.024, "math_500": 0.121, "aime": 0, "aime_25": null, "ifbench": 0.199, "lcr": 0, "terminalbench_hard": null, "tau2": 0 } }, { "id": "mistral-small-3", "name": "Mistral: Mistral Small 3", "short_name": "Mistral Small 3", "description": "Mistral Small 3 is competitive with larger models such as Llama 3.3 70B or Qwen 32B, and is an excellent open replacement for opaque proprietary models like GPT4o-mini. Mistral Small 3 is on par with Llama 3.3 70B instruct", "context_length": 32000, "created": 1738195200, "owned_by": "mistral", "available": true, "hallucination_rate": 3.1, "endpoints": [ "/v1/chat/completions" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.14 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.28 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0 } } }, "supported_parameters": [ "frequency_penalty", "logit_bias", "max_completion_tokens", "presence_penalty", "reasoning_effort", "response_format", "stop", "temperature", "tool_choice", "tools", "top_p" ], "capabilities": { "vision": false, "function_call": false, "reasoning": false }, "benchmarks": { "artificial_analysis_intelligence_index": 12.7, "artificial_analysis_coding_index": null, "artificial_analysis_math_index": 4.3, "mmlu_pro": 0.652, "gpqa": 0.462, "hle": 0.041, "livecodebench": 0.252, "scicode": 0.236, "math_500": 0.715, "aime": 0.08, "aime_25": 0.043, "ifbench": 0.264, "lcr": 0, "terminalbench_hard": null, "tau2": 0.196 } }, { "id": "deepseek-v3.1-nex-n1", "name": "Nex AGI: DeepSeek V3.1 Nex N1", "short_name": "DeepSeek V3.1 Nex N1", "description": "DeepSeek V3.1 Nex-N1 is the flagship release of the Nex-N1 series — a post-trained model designed to highlight agent autonomy, tool use, and real-world productivity. \n\nNex-N1 demonstrates competitive performance across all evaluation scenarios, showing particularly strong results in practical coding and HTML generation tasks.", "context_length": 131072, "created": 1765204393, "owned_by": "nex-agi", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/chat/completions", "/v1/responses", "/v1/messages" ], "pricing": { "input": { "unit": "1M tokens", "price_per_million_t": 0.25 }, "output": { "unit": "1M tokens", "price_per_million_t": 0.39 }, "caching": { "read": { "unit": "1M tokens", "price_per_million_t": 0 } } }, "supported_parameters": [ "frequency_penalty", "response_format", "structured_outputs", "temperature", "tool_choice", "tools", "top_k", "top_p" ], "capabilities": { "vision": false, "function_call": true, "reasoning": false }, "benchmarks": { "artificial_analysis_intelligence_index": null, "artificial_analysis_coding_index": null, "artificial_analysis_math_index": null, "mmlu_pro": null, "gpqa": null, "hle": null, "livecodebench": null, "scicode": null, "math_500": null, "aime": null, "aime_25": null, "ifbench": null, "lcr": null, "terminalbench_hard": null, "tau2": null }, "outdated": true }, { "id": "gpt-image-1.5", "name": "OpenAI: GPT Image 1.5", "short_name": "GPT Image 1.5", "description": "OpenAI's latest image model with better instruction following and adherence to prompts. Up to 4x faster than its predecessor with improved text rendering and iterative editing capabilities.", "created": 1765929600, "owned_by": "openai", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/images/generations", "/v1/images/edits" ], "pricing": { "unit": "per image", "resolutions": { "1024x1024": 0.1666, "1024x1536": 0.2499, "1536x1024": 0.2499, "auto": 0.1666 } }, "supported_sizes": [ "1024x1024", "1536x1024", "1024x1536" ], "benchmarks": { "elo": 1266, "rank": 1, "ci95": "-11/11", "appearances": 4654, "categories": { "style_general_photorealistic": { "elo": 1274, "ci95": "-18/18", "appearances": 1719 }, "style_cartoon_illustration": { "elo": 1277, "ci95": "-40/40", "appearances": 403 }, "subject_people:_portraits": { "elo": 1274, "ci95": "-35/35", "appearances": 523 }, "style_anime": { "elo": 1258, "ci95": "-43/43", "appearances": 346 }, "subject_futuristic_sci-fi": { "elo": 1265, "ci95": "-32/32", "appearances": 602 }, "subject_physical_spaces": { "elo": 1234, "ci95": "-32/32", "appearances": 565 }, "subject_text_typography": { "elo": 1210, "ci95": "-58/58", "appearances": 177 }, "subject_nature_landscapes": { "elo": 1262, "ci95": "-36/36", "appearances": 435 }, "style_vintage_retro": { "elo": 1315, "ci95": "-57/57", "appearances": 227 }, "subject_fantasy_mythical": { "elo": 1325, "ci95": "-33/33", "appearances": 615 }, "subject_ui/ux_design": { "elo": 1264, "ci95": "-66/66", "appearances": 177 }, "subject_commercial": { "elo": 1202, "ci95": "-41/41", "appearances": 326 }, "style_traditional_art": { "elo": 1366, "ci95": "-50/50", "appearances": 275 }, "style_graphic_design_digital_rendering": { "elo": 1234, "ci95": "-21/21", "appearances": 1428 }, "subject_people:_groups_activities": { "elo": 1334, "ci95": "-38/38", "appearances": 470 } } }, "discount": { "percentage": 5, "note": "A 5% discount applies to input and output token pricing only." } }, { "id": "nano-banana-2", "name": "Nano Banana 2", "short_name": "Nano Banana 2", "description": "Google Nano Banana 2 (Gemini 3.1 Flash Image) for fast text-to-image", "created": 1772064000, "owned_by": "google", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/images/generations", "/v1/images/edits" ], "pricing": { "unit": "per image", "resolutions": { "1k": 0.08, "2k": 0.12, "4k": 0.16, "auto": 0.08 } }, "supported_sizes": [ "1k", "2k", "4k" ], "benchmarks": { "elo": 1258, "rank": 2, "ci95": "-9/9", "appearances": 5912, "categories": { "style_general_photorealistic": { "elo": 1296, "ci95": "-18/18", "appearances": 1841 }, "style_cartoon_illustration": { "elo": 1289, "ci95": "-36/36", "appearances": 494 }, "subject_people:_portraits": { "elo": 1280, "ci95": "-34/34", "appearances": 548 }, "style_anime": { "elo": 1181, "ci95": "-35/35", "appearances": 424 }, "subject_futuristic_sci-fi": { "elo": 1224, "ci95": "-28/28", "appearances": 717 }, "subject_physical_spaces": { "elo": 1255, "ci95": "-30/30", "appearances": 666 }, "subject_text_typography": { "elo": 1190, "ci95": "-40/40", "appearances": 331 }, "subject_nature_landscapes": { "elo": 1255, "ci95": "-35/35", "appearances": 463 }, "style_vintage_retro": { "elo": 1259, "ci95": "-41/41", "appearances": 336 }, "subject_fantasy_mythical": { "elo": 1260, "ci95": "-26/26", "appearances": 768 }, "subject_ui/ux_design": { "elo": 1196, "ci95": "-54/54", "appearances": 211 }, "subject_commercial": { "elo": 1258, "ci95": "-38/38", "appearances": 386 }, "style_traditional_art": { "elo": 1286, "ci95": "-42/42", "appearances": 294 }, "style_graphic_design_digital_rendering": { "elo": 1211, "ci95": "-18/18", "appearances": 1691 }, "subject_people:_groups_activities": { "elo": 1346, "ci95": "-35/35", "appearances": 540 } } }, "discount": { "percentage": 5, "note": "A 5% discount applies to input and output token pricing only." } }, { "id": "nano-banana-2-fast", "name": "Nano Banana 2 Fast", "short_name": "Nano Banana 2 Fast", "description": "Google Nano Banana 2 Fast (Gemini 3.1 Flash Image) is the cheaper Nano Banana 2 variant for unified text-to-image", "created": 1772755200, "owned_by": "google", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/images/generations", "/v1/images/edits" ], "pricing": { "unit": "per image", "resolutions": { "2k": 0.045, "4k": 0.075, "auto": 0.045 } }, "supported_sizes": [ "2k", "4k" ], "benchmarks": null, "discount": { "percentage": 5, "note": "A 5% discount applies to input and output token pricing only." } }, { "id": "nano-banana-pro", "name": "Nano Banana Pro", "short_name": "Nano Banana Pro", "description": "Nano Banana 2 is Google's latest AI image generation model from DeepMind. It combines the advanced capabilities of Nano Banana Pro — including high-quality outputs, strong subject consistency, precise prompt following, and rich world knowledge — with the lightning-fast speed of Gemini Flash.", "created": 1763683200, "owned_by": "google", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/images/generations", "/v1/images/edits" ], "pricing": { "unit": "per image", "resolutions": { "1k": 0.14, "2k": 0.14, "4k": 0.24, "auto": 0.14 } }, "supported_sizes": [ "1k", "2k", "4k" ], "benchmarks": { "elo": 1214, "rank": 4, "ci95": "-11/11", "appearances": 3897, "categories": { "style_general_photorealistic": { "elo": 1255, "ci95": "-20/20", "appearances": 1348 }, "style_cartoon_illustration": { "elo": 1240, "ci95": "-43/43", "appearances": 304 }, "subject_people:_portraits": { "elo": 1188, "ci95": "-36/36", "appearances": 401 }, "style_anime": { "elo": 1153, "ci95": "-43/43", "appearances": 283 }, "subject_futuristic_sci-fi": { "elo": 1193, "ci95": "-31/31", "appearances": 537 }, "subject_physical_spaces": { "elo": 1201, "ci95": "-33/33", "appearances": 496 }, "subject_text_typography": { "elo": 1188, "ci95": "-55/55", "appearances": 177 }, "subject_nature_landscapes": { "elo": 1222, "ci95": "-39/39", "appearances": 337 }, "style_vintage_retro": { "elo": 1252, "ci95": "-54/54", "appearances": 212 }, "subject_fantasy_mythical": { "elo": 1228, "ci95": "-32/32", "appearances": 497 }, "subject_ui/ux_design": { "elo": 1142, "ci95": "-64/64", "appearances": 145 }, "subject_commercial": { "elo": 1230, "ci95": "-44/44", "appearances": 288 }, "style_traditional_art": { "elo": 1254, "ci95": "-47/47", "appearances": 237 }, "style_graphic_design_digital_rendering": { "elo": 1163, "ci95": "-20/20", "appearances": 1276 }, "subject_people:_groups_activities": { "elo": 1322, "ci95": "-41/41", "appearances": 368 } } }, "discount": { "percentage": 5, "note": "A 5% discount applies to input and output token pricing only." } }, { "id": "nano-banana-pro-ultra", "name": "Nano Banana Pro Ultra", "short_name": "Nano Banana Pro Ultra", "description": "Google's Nano Banana Pro Ultra (Gemini 3.0 Pro Image) pushes our phone-optimized pipeline to 4K and 8K detail. It's tuned for instant, high-clarity compositions, balanced lighting, and accurate scene understanding straight from natural language prompts.", "created": 1763683200, "owned_by": "google", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/images/generations", "/v1/images/edits" ], "pricing": { "unit": "per image", "resolutions": { "4k": 0.15, "8k": 0.18, "auto": 0.15 } }, "supported_sizes": [ "4k", "8k" ], "benchmarks": { "elo": 1214, "rank": 4, "ci95": "-11/11", "appearances": 3897, "categories": { "style_general_photorealistic": { "elo": 1255, "ci95": "-20/20", "appearances": 1348 }, "style_cartoon_illustration": { "elo": 1240, "ci95": "-43/43", "appearances": 304 }, "subject_people:_portraits": { "elo": 1188, "ci95": "-36/36", "appearances": 401 }, "style_anime": { "elo": 1153, "ci95": "-43/43", "appearances": 283 }, "subject_futuristic_sci-fi": { "elo": 1193, "ci95": "-31/31", "appearances": 537 }, "subject_physical_spaces": { "elo": 1201, "ci95": "-33/33", "appearances": 496 }, "subject_text_typography": { "elo": 1188, "ci95": "-55/55", "appearances": 177 }, "subject_nature_landscapes": { "elo": 1222, "ci95": "-39/39", "appearances": 337 }, "style_vintage_retro": { "elo": 1252, "ci95": "-54/54", "appearances": 212 }, "subject_fantasy_mythical": { "elo": 1228, "ci95": "-32/32", "appearances": 497 }, "subject_ui/ux_design": { "elo": 1142, "ci95": "-64/64", "appearances": 145 }, "subject_commercial": { "elo": 1230, "ci95": "-44/44", "appearances": 288 }, "style_traditional_art": { "elo": 1254, "ci95": "-47/47", "appearances": 237 }, "style_graphic_design_digital_rendering": { "elo": 1163, "ci95": "-20/20", "appearances": 1276 }, "subject_people:_groups_activities": { "elo": 1322, "ci95": "-41/41", "appearances": 368 } } }, "discount": { "percentage": 5, "note": "A 5% discount applies to input and output token pricing only." } }, { "id": "seedream-v4", "name": "ByteDance: Seedream 4.0", "short_name": "Seedream 4.0", "description": "Seedream 4.0 is a state-of-art image model.", "created": 1757203200, "owned_by": "bytedance", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/images/generations", "/v1/images/edits" ], "pricing": { "unit": "per image", "resolutions": { "1024x1024": 0.033, "1536x1024": 0.033, "1024x1536": 0.033, "2048x2048": 0.033, "4096x4096": 0.033, "auto": 0.033 } }, "supported_sizes": [ "2048x2048", "1024x1024", "1536x1024", "1024x1536", "3072x2048", "2048x3072", "4096x2304", "2304x4096", "4096x4096" ], "benchmarks": { "elo": 1185, "rank": 6, "ci95": "-7/7", "appearances": 9957, "categories": { "style_general_photorealistic": { "elo": 1222, "ci95": "-11/11", "appearances": 3943 }, "style_cartoon_illustration": { "elo": 1173, "ci95": "-24/24", "appearances": 820 }, "subject_people:_portraits": { "elo": 1187, "ci95": "-20/20", "appearances": 1191 }, "style_anime": { "elo": 1199, "ci95": "-24/24", "appearances": 835 }, "subject_futuristic_sci-fi": { "elo": 1181, "ci95": "-19/19", "appearances": 1322 }, "subject_physical_spaces": { "elo": 1184, "ci95": "-19/19", "appearances": 1281 }, "subject_text_typography": { "elo": 1108, "ci95": "-31/31", "appearances": 482 }, "subject_nature_landscapes": { "elo": 1187, "ci95": "-22/22", "appearances": 947 }, "style_vintage_retro": { "elo": 1200, "ci95": "-30/30", "appearances": 529 }, "subject_fantasy_mythical": { "elo": 1226, "ci95": "-18/18", "appearances": 1474 }, "subject_ui/ux_design": { "elo": 1063, "ci95": "-35/35", "appearances": 393 }, "subject_commercial": { "elo": 1137, "ci95": "-27/27", "appearances": 667 }, "style_traditional_art": { "elo": 1237, "ci95": "-27/27", "appearances": 661 }, "style_graphic_design_digital_rendering": { "elo": 1132, "ci95": "-13/13", "appearances": 2834 }, "subject_people:_groups_activities": { "elo": 1265, "ci95": "-21/21", "appearances": 1107 } } } }, { "id": "seedream-v4.5", "name": "ByteDance: Seedream 4.5", "short_name": "Seedream 4.5", "description": "Seedream 4.5 is the latest image model with improved quality.", "created": 1764720000, "owned_by": "bytedance", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/images/generations", "/v1/images/edits" ], "pricing": { "unit": "per image", "resolutions": { "1920x1920": 0.044, "2048x2048": 0.044, "2496x1664": 0.044, "1664x2496": 0.044, "3072x2048": 0.044, "2048x3072": 0.044, "4096x2304": 0.044, "2304x4096": 0.044, "4096x4096": 0.044, "auto": 0.044 } }, "supported_sizes": [ "1920x1920", "2048x2048", "2496x1664", "1664x2496", "3072x2048", "2048x3072", "4096x2304", "2304x4096", "4096x4096" ], "benchmarks": { "elo": 1172, "rank": 10, "ci95": "-9/9", "appearances": 5785, "categories": { "style_general_photorealistic": { "elo": 1175, "ci95": "-15/15", "appearances": 2188 }, "style_cartoon_illustration": { "elo": 1236, "ci95": "-34/34", "appearances": 504 }, "subject_people:_portraits": { "elo": 1186, "ci95": "-30/30", "appearances": 600 }, "style_anime": { "elo": 1209, "ci95": "-35/35", "appearances": 475 }, "subject_futuristic_sci-fi": { "elo": 1182, "ci95": "-26/26", "appearances": 768 }, "subject_physical_spaces": { "elo": 1118, "ci95": "-24/24", "appearances": 819 }, "subject_text_typography": { "elo": 1076, "ci95": "-47/47", "appearances": 231 }, "subject_nature_landscapes": { "elo": 1176, "ci95": "-31/31", "appearances": 529 }, "style_vintage_retro": { "elo": 1173, "ci95": "-45/45", "appearances": 268 }, "subject_fantasy_mythical": { "elo": 1244, "ci95": "-25/25", "appearances": 882 }, "subject_ui/ux_design": { "elo": 1129, "ci95": "-56/56", "appearances": 187 }, "subject_commercial": { "elo": 1163, "ci95": "-37/37", "appearances": 399 }, "style_traditional_art": { "elo": 1251, "ci95": "-38/38", "appearances": 369 }, "style_graphic_design_digital_rendering": { "elo": 1133, "ci95": "-17/17", "appearances": 1714 }, "subject_people:_groups_activities": { "elo": 1226, "ci95": "-29/29", "appearances": 628 } } } }, { "id": "seedream-v4.5-sequential", "name": "ByteDance: Seedream 4.5 Sequential", "short_name": "Seedream 4.5 Sequential", "description": "Seedream 4.5 Sequential generates multiple consistent images with character and object consistency. Maintains unified palette, lighting, and style across outputs.", "created": 1764720000, "owned_by": "bytedance", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/images/generations", "/v1/images/edits" ], "pricing": { "unit": "per image", "resolutions": { "2560x1440": 0.044, "1440x2560": 0.044, "1920x1920": 0.044, "3072x2048": 0.044, "2048x3072": 0.044, "4096x2304": 0.044, "2304x4096": 0.044, "4096x4096": 0.044, "auto": 0.044 } }, "supported_sizes": [ "2560x1440", "1440x2560", "1920x1920", "3072x2048", "2048x3072", "4096x2304", "2304x4096", "4096x4096" ], "benchmarks": null }, { "id": "seedream-v5.0-lite-sequential", "name": "ByteDance: Seedream 5.0 Lite Sequential", "short_name": "Seedream 5.0 Lite Sequential", "description": "Seedream 5.0 Lite Sequential generates consistent multi-image sets in one run, with unified style, palette, and character continuity.", "created": 1771891200, "owned_by": "bytedance", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/images/generations", "/v1/images/edits" ], "pricing": { "unit": "per image", "resolutions": { "2560x1440": 0.0385, "1440x2560": 0.0385, "1920x1920": 0.0385, "3072x2048": 0.0385, "2048x3072": 0.0385, "4096x2304": 0.0385, "2304x4096": 0.0385, "auto": 0.0385 } }, "supported_sizes": [ "2560x1440", "1440x2560", "1920x1920", "3072x2048", "2048x3072", "4096x2304", "2304x4096" ], "benchmarks": { "elo": 1122, "rank": 30, "ci95": "-11/11", "appearances": 3508, "categories": { "style_general_photorealistic": { "elo": 1107, "ci95": "-21/21", "appearances": 1124 }, "style_cartoon_illustration": { "elo": 1198, "ci95": "-42/42", "appearances": 294 }, "subject_people:_portraits": { "elo": 1159, "ci95": "-38/38", "appearances": 357 }, "style_anime": { "elo": 1139, "ci95": "-45/45", "appearances": 243 }, "subject_futuristic_sci-fi": { "elo": 1072, "ci95": "-34/34", "appearances": 418 }, "subject_physical_spaces": { "elo": 1085, "ci95": "-34/34", "appearances": 411 }, "subject_text_typography": { "elo": 1105, "ci95": "-54/54", "appearances": 169 }, "subject_nature_landscapes": { "elo": 1100, "ci95": "-42/42", "appearances": 275 }, "style_vintage_retro": { "elo": 1106, "ci95": "-56/56", "appearances": 158 }, "subject_fantasy_mythical": { "elo": 1175, "ci95": "-32/32", "appearances": 466 }, "subject_commercial": { "elo": 1136, "ci95": "-45/45", "appearances": 243 }, "style_traditional_art": { "elo": 1182, "ci95": "-48/48", "appearances": 207 }, "style_graphic_design_digital_rendering": { "elo": 1093, "ci95": "-22/22", "appearances": 996 }, "subject_people:_groups_activities": { "elo": 1152, "ci95": "-39/39", "appearances": 324 } } } }, { "id": "hunyuan-image-3", "name": "Tencent: Hunyuan Image 3", "short_name": "Hunyuan Image 3", "description": "State-of-the-art text-to-image model producing high-quality, emotionally resonant visuals with strong prompt adherence.", "created": 1760054400, "owned_by": "tencent", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/images/generations", "/v1/images/edits" ], "pricing": { "unit": "per image", "resolutions": { "auto": 0.055, "512x512": 0.055, "768x1024": 0.055, "1024x768": 0.055, "1024x1024": 0.055, "1024x1536": 0.055, "1536x1024": 0.055 } }, "supported_sizes": [ "1024x1024", "768x1024", "1024x768", "1024x1536", "1536x1024", "512x512" ], "benchmarks": null }, { "id": "hunyuan-image-3-instruct", "name": "Tencent: Hunyuan Image 3 Instruct", "short_name": "Hunyuan Image 3 Instruct", "description": "Tencent's Hunyuan Image 3 Instruct model for text-to-image generation.", "created": 1769644800, "owned_by": "tencent", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/images/generations", "/v1/images/edits" ], "pricing": { "unit": "per image", "resolutions": { "1024x1024": 0.132, "1024x1536": 0.132, "1536x1024": 0.132, "768x1024": 0.132, "1024x768": 0.132, "512x512": 0.132, "256x256": 0.132, "auto": 0.132 } }, "supported_sizes": [ "1024x1024", "1024x1536", "1536x1024", "768x1024", "1024x768", "512x512", "256x256" ], "benchmarks": null }, { "id": "qwen-image-2512", "name": "Qwen: Qwen Image 2512", "short_name": "Qwen Image 2512", "description": "Alibaba Qwen's latest text-to-image model with strong prompt understanding and high-fidelity text rendering. Supports flexible sizing and reproducible seeds.", "created": 1766361600, "owned_by": "qwen", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/images/generations", "/v1/images/edits" ], "pricing": { "unit": "per image", "resolutions": { "1024x1024": 0.0275, "1024x1536": 0.0275, "1536x1024": 0.0275, "1536x1536": 0.0275, "768x1024": 0.0275, "1024x768": 0.0275, "512x512": 0.0275, "256x256": 0.0275, "auto": 0.0275 } }, "supported_sizes": [ "1024x1024", "1024x1536", "1536x1024", "1536x1536", "768x1024", "1024x768", "512x512", "256x256" ], "benchmarks": { "elo": 1151, "rank": 16, "ci95": "-12/12", "appearances": 3172, "categories": { "style_general_photorealistic": { "elo": 1191, "ci95": "-20/20", "appearances": 1278 }, "style_cartoon_illustration": { "elo": 1171, "ci95": "-42/42", "appearances": 297 }, "subject_people:_portraits": { "elo": 1133, "ci95": "-36/36", "appearances": 390 }, "style_anime": { "elo": 1109, "ci95": "-45/45", "appearances": 251 }, "subject_futuristic_sci-fi": { "elo": 1155, "ci95": "-35/35", "appearances": 419 }, "subject_physical_spaces": { "elo": 1132, "ci95": "-34/34", "appearances": 419 }, "subject_nature_landscapes": { "elo": 1179, "ci95": "-40/40", "appearances": 312 }, "subject_fantasy_mythical": { "elo": 1169, "ci95": "-31/31", "appearances": 525 }, "subject_commercial": { "elo": 1148, "ci95": "-55/55", "appearances": 181 }, "style_traditional_art": { "elo": 1162, "ci95": "-50/50", "appearances": 193 }, "style_graphic_design_digital_rendering": { "elo": 1120, "ci95": "-24/24", "appearances": 880 }, "subject_people:_groups_activities": { "elo": 1202, "ci95": "-39/39", "appearances": 346 } } } }, { "id": "qwen-image-max", "name": "Qwen: Qwen Image Max", "short_name": "Qwen Image Max", "description": "Qwen Image Max is a text-to-image model with high-quality image generation supporting Chinese and English prompts.", "created": 1769644800, "owned_by": "qwen", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/images/generations", "/v1/images/edits" ], "pricing": { "unit": "per image", "resolutions": { "1024x1024": 0.077, "1280x720": 0.077, "720x1280": 0.077, "1536x1024": 0.077, "1024x1536": 0.077, "auto": 0.077 } }, "supported_sizes": [ "1024x1024", "1280x720", "720x1280", "1536x1024", "1024x1536" ], "benchmarks": { "elo": 1151, "rank": 16, "ci95": "-12/12", "appearances": 3172, "categories": { "style_general_photorealistic": { "elo": 1191, "ci95": "-20/20", "appearances": 1278 }, "style_cartoon_illustration": { "elo": 1171, "ci95": "-42/42", "appearances": 297 }, "subject_people:_portraits": { "elo": 1133, "ci95": "-36/36", "appearances": 390 }, "style_anime": { "elo": 1109, "ci95": "-45/45", "appearances": 251 }, "subject_futuristic_sci-fi": { "elo": 1155, "ci95": "-35/35", "appearances": 419 }, "subject_physical_spaces": { "elo": 1132, "ci95": "-34/34", "appearances": 419 }, "subject_nature_landscapes": { "elo": 1179, "ci95": "-40/40", "appearances": 312 }, "subject_fantasy_mythical": { "elo": 1169, "ci95": "-31/31", "appearances": 525 }, "subject_commercial": { "elo": 1148, "ci95": "-55/55", "appearances": 181 }, "style_traditional_art": { "elo": 1162, "ci95": "-50/50", "appearances": 193 }, "style_graphic_design_digital_rendering": { "elo": 1120, "ci95": "-24/24", "appearances": 880 }, "subject_people:_groups_activities": { "elo": 1202, "ci95": "-39/39", "appearances": 346 } } } }, { "id": "cogview-4", "name": "Z.ai: CogView-4", "short_name": "CogView-4", "description": "CogView-4 turns detailed prompts into precise, high-quality images with strong composition and quick iteration. Supports standard (fast) and HD (richer detail) quality modes plus multiple aspect ratios.", "created": 1767657600, "owned_by": "z-ai", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/images/generations", "/v1/images/edits" ], "pricing": { "unit": "per image", "resolutions": { "1024x1024": 0.011, "768x1344": 0.011, "864x1152": 0.011, "1344x768": 0.011, "1152x864": 0.011, "1440x720": 0.011, "720x1440": 0.011, "auto": 0.011 } }, "supported_sizes": [ "1024x1024", "768x1344", "864x1152", "1344x768", "1152x864", "1440x720", "720x1440" ], "benchmarks": null }, { "id": "glm-image", "name": "Z.ai: GLM Image", "short_name": "GLM Image", "description": "GLM Image is a text-to-image model with flexible sizing, prompt expansion, and JPEG/PNG/WebP output formats.", "created": 1768348800, "owned_by": "z-ai", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/images/generations", "/v1/images/edits" ], "pricing": { "unit": "per image", "resolutions": { "1024x1024": 0.02805, "1024x1536": 0.02805, "1536x1024": 0.02805, "768x1024": 0.02805, "1024x768": 0.02805, "512x512": 0.02805, "256x256": 0.02805, "auto": 0.02805 } }, "supported_sizes": [ "1024x1024", "1024x1536", "1536x1024", "768x1024", "1024x768", "512x512", "256x256" ], "benchmarks": null }, { "id": "nano-banana", "name": "Google: Nano Banana", "short_name": "Nano Banana", "description": "Google's lightweight text-to-image model. Fast, high-quality visuals with versatile style support.", "created": 1704067200, "owned_by": "google", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/images/generations", "/v1/images/edits" ], "pricing": { "unit": "per image", "resolutions": { "1k": 0.08, "2k": 0.12, "4k": 0.16, "auto": 0.08 } }, "supported_sizes": [ "1k", "2k", "4k" ], "benchmarks": { "elo": 1164, "rank": 15, "ci95": "-7/7", "appearances": 8883, "categories": { "style_general_photorealistic": { "elo": 1188, "ci95": "-11/11", "appearances": 3604 }, "style_cartoon_illustration": { "elo": 1193, "ci95": "-26/26", "appearances": 727 }, "subject_people:_portraits": { "elo": 1178, "ci95": "-21/21", "appearances": 1075 }, "style_anime": { "elo": 1144, "ci95": "-25/25", "appearances": 775 }, "subject_futuristic_sci-fi": { "elo": 1137, "ci95": "-20/20", "appearances": 1188 }, "subject_physical_spaces": { "elo": 1157, "ci95": "-20/20", "appearances": 1181 }, "subject_text_typography": { "elo": 1131, "ci95": "-34/34", "appearances": 411 }, "subject_nature_landscapes": { "elo": 1158, "ci95": "-23/23", "appearances": 873 }, "style_vintage_retro": { "elo": 1153, "ci95": "-32/32", "appearances": 467 }, "subject_fantasy_mythical": { "elo": 1204, "ci95": "-19/19", "appearances": 1348 }, "subject_ui/ux_design": { "elo": 1100, "ci95": "-41/41", "appearances": 289 }, "subject_commercial": { "elo": 1133, "ci95": "-29/29", "appearances": 558 }, "style_traditional_art": { "elo": 1245, "ci95": "-27/27", "appearances": 629 }, "style_graphic_design_digital_rendering": { "elo": 1125, "ci95": "-14/14", "appearances": 2413 }, "subject_people:_groups_activities": { "elo": 1209, "ci95": "-22/22", "appearances": 993 } } }, "discount": { "percentage": 5, "note": "A 5% discount applies to input and output token pricing only." } }, { "id": "riverflow-2-fast", "name": "Sourceful: Riverflow 2 Fast", "short_name": "Riverflow 2 Fast", "description": "Lightweight Riverflow 2 variant tuned for quick generations, ideal for fast iterations on packaging and layouts.", "created": 1704067200, "owned_by": "sourceful", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/images/generations", "/v1/images/edits" ], "pricing": { "unit": "per image", "resolutions": { "1024x1024": 0.033, "auto": 0.033 } }, "supported_sizes": [ "1024x1024" ], "benchmarks": { "elo": 1254, "rank": 3, "ci95": "-11/11", "appearances": 4833, "categories": { "style_general_photorealistic": { "elo": 1291, "ci95": "-18/18", "appearances": 1951 }, "style_cartoon_illustration": { "elo": 1260, "ci95": "-38/38", "appearances": 418 }, "subject_people:_portraits": { "elo": 1280, "ci95": "-32/32", "appearances": 607 }, "style_anime": { "elo": 1206, "ci95": "-37/37", "appearances": 409 }, "subject_futuristic_sci-fi": { "elo": 1237, "ci95": "-30/30", "appearances": 638 }, "subject_physical_spaces": { "elo": 1246, "ci95": "-30/30", "appearances": 672 }, "subject_text_typography": { "elo": 1163, "ci95": "-51/51", "appearances": 207 }, "subject_nature_landscapes": { "elo": 1253, "ci95": "-33/33", "appearances": 527 }, "style_vintage_retro": { "elo": 1274, "ci95": "-56/56", "appearances": 201 }, "subject_fantasy_mythical": { "elo": 1275, "ci95": "-27/27", "appearances": 764 }, "subject_ui/ux_design": { "elo": 1187, "ci95": "-72/72", "appearances": 126 }, "subject_commercial": { "elo": 1288, "ci95": "-51/51", "appearances": 248 }, "style_traditional_art": { "elo": 1326, "ci95": "-42/42", "appearances": 341 }, "style_graphic_design_digital_rendering": { "elo": 1204, "ci95": "-20/20", "appearances": 1349 }, "subject_people:_groups_activities": { "elo": 1322, "ci95": "-34/34", "appearances": 564 } } } }, { "id": "riverflow-2-standard", "name": "Sourceful: Riverflow 2 Standard", "short_name": "Riverflow 2 Standard", "description": "Balanced Riverflow 2 variant combining realistic output, strong prompt adherence, and robust editing performance.", "created": 1704067200, "owned_by": "sourceful", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/images/generations", "/v1/images/edits" ], "pricing": { "unit": "per image", "resolutions": { "1024x1024": 0.0385, "auto": 0.0385 } }, "supported_sizes": [ "1024x1024" ], "benchmarks": { "elo": 1254, "rank": 3, "ci95": "-11/11", "appearances": 4833, "categories": { "style_general_photorealistic": { "elo": 1291, "ci95": "-18/18", "appearances": 1951 }, "style_cartoon_illustration": { "elo": 1260, "ci95": "-38/38", "appearances": 418 }, "subject_people:_portraits": { "elo": 1280, "ci95": "-32/32", "appearances": 607 }, "style_anime": { "elo": 1206, "ci95": "-37/37", "appearances": 409 }, "subject_futuristic_sci-fi": { "elo": 1237, "ci95": "-30/30", "appearances": 638 }, "subject_physical_spaces": { "elo": 1246, "ci95": "-30/30", "appearances": 672 }, "subject_text_typography": { "elo": 1163, "ci95": "-51/51", "appearances": 207 }, "subject_nature_landscapes": { "elo": 1253, "ci95": "-33/33", "appearances": 527 }, "style_vintage_retro": { "elo": 1274, "ci95": "-56/56", "appearances": 201 }, "subject_fantasy_mythical": { "elo": 1275, "ci95": "-27/27", "appearances": 764 }, "subject_ui/ux_design": { "elo": 1187, "ci95": "-72/72", "appearances": 126 }, "subject_commercial": { "elo": 1288, "ci95": "-51/51", "appearances": 248 }, "style_traditional_art": { "elo": 1326, "ci95": "-42/42", "appearances": 341 }, "style_graphic_design_digital_rendering": { "elo": 1204, "ci95": "-20/20", "appearances": 1349 }, "subject_people:_groups_activities": { "elo": 1322, "ci95": "-34/34", "appearances": 564 } } } }, { "id": "riverflow-2-max", "name": "Sourceful: Riverflow 2 Max", "short_name": "Riverflow 2 Max", "description": "Highest-fidelity Riverflow 2 variant for production-ready creative with the strongest detail and editing quality.", "created": 1704067200, "owned_by": "sourceful", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/images/generations", "/v1/images/edits" ], "pricing": { "unit": "per image", "resolutions": { "1024x1024": 0.0825, "auto": 0.0825 } }, "supported_sizes": [ "1024x1024" ], "benchmarks": { "elo": 1254, "rank": 3, "ci95": "-11/11", "appearances": 4833, "categories": { "style_general_photorealistic": { "elo": 1291, "ci95": "-18/18", "appearances": 1951 }, "style_cartoon_illustration": { "elo": 1260, "ci95": "-38/38", "appearances": 418 }, "subject_people:_portraits": { "elo": 1280, "ci95": "-32/32", "appearances": 607 }, "style_anime": { "elo": 1206, "ci95": "-37/37", "appearances": 409 }, "subject_futuristic_sci-fi": { "elo": 1237, "ci95": "-30/30", "appearances": 638 }, "subject_physical_spaces": { "elo": 1246, "ci95": "-30/30", "appearances": 672 }, "subject_text_typography": { "elo": 1163, "ci95": "-51/51", "appearances": 207 }, "subject_nature_landscapes": { "elo": 1253, "ci95": "-33/33", "appearances": 527 }, "style_vintage_retro": { "elo": 1274, "ci95": "-56/56", "appearances": 201 }, "subject_fantasy_mythical": { "elo": 1275, "ci95": "-27/27", "appearances": 764 }, "subject_ui/ux_design": { "elo": 1187, "ci95": "-72/72", "appearances": 126 }, "subject_commercial": { "elo": 1288, "ci95": "-51/51", "appearances": 248 }, "style_traditional_art": { "elo": 1326, "ci95": "-42/42", "appearances": 341 }, "style_graphic_design_digital_rendering": { "elo": 1204, "ci95": "-20/20", "appearances": 1349 }, "subject_people:_groups_activities": { "elo": 1322, "ci95": "-34/34", "appearances": 564 } } } }, { "id": "gpt-image-1", "name": "OpenAI: GPT 4o Image", "short_name": "GPT 4o Image", "description": "OpenAI's latest image model supporting both text-to-image.", "created": 1704067200, "owned_by": "openai", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/images/generations", "/v1/images/edits" ], "pricing": { "unit": "per image", "resolutions": { "1024x1024": 0.2839, "1024x1536": 0.425, "1536x1024": 0.425, "auto": 0.2839 } }, "supported_sizes": [ "1024x1024", "1536x1024", "1024x1536" ], "benchmarks": null, "discount": { "percentage": 5, "note": "A 5% discount applies to input and output token pricing only." } }, { "id": "gpt-image-1-mini", "name": "OpenAI: GPT Image 1 Mini", "short_name": "GPT Image 1 Mini", "description": "Cost-efficient OpenAI image model via Wavespeed.", "created": 1761264000, "owned_by": "openai", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/images/generations", "/v1/images/edits" ], "pricing": { "unit": "per image", "resolutions": { "1024x1024": 0.02, "auto": 0.02 } }, "supported_sizes": [ "1024x1024" ], "benchmarks": { "elo": 1072, "rank": 53, "ci95": "-11/11", "appearances": 3784, "categories": { "style_general_photorealistic": { "elo": 1083, "ci95": "-18/18", "appearances": 1516 }, "style_cartoon_illustration": { "elo": 1068, "ci95": "-38/38", "appearances": 331 }, "subject_people:_portraits": { "elo": 1112, "ci95": "-33/33", "appearances": 439 }, "style_anime": { "elo": 1026, "ci95": "-42/42", "appearances": 283 }, "subject_futuristic_sci-fi": { "elo": 1053, "ci95": "-31/31", "appearances": 503 }, "subject_physical_spaces": { "elo": 1040, "ci95": "-32/32", "appearances": 463 }, "subject_text_typography": { "elo": 1070, "ci95": "-61/61", "appearances": 134 }, "subject_nature_landscapes": { "elo": 1123, "ci95": "-34/34", "appearances": 415 }, "style_vintage_retro": { "elo": 1140, "ci95": "-53/53", "appearances": 174 }, "subject_fantasy_mythical": { "elo": 1074, "ci95": "-29/29", "appearances": 574 }, "subject_commercial": { "elo": 1062, "ci95": "-48/48", "appearances": 208 }, "style_traditional_art": { "elo": 1226, "ci95": "-43/43", "appearances": 261 }, "style_graphic_design_digital_rendering": { "elo": 1018, "ci95": "-21/21", "appearances": 1045 }, "subject_people:_groups_activities": { "elo": 1078, "ci95": "-34/34", "appearances": 438 } } }, "discount": { "percentage": 5, "note": "A 5% discount applies to input and output token pricing only." } }, { "id": "riverflow-2.0-pro", "name": "Sourceful: Riverflow 2.0 Pro", "short_name": "Riverflow 2.0 Pro", "description": "Agentic image model optimized for robust, high-precision text-to-image generation.", "created": 1770163200, "owned_by": "sourceful", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/images/generations", "/v1/images/edits" ], "pricing": { "unit": "per image", "resolutions": { "1k": 0.1485, "2k": 0.1485, "4k": 0.3267, "auto": 0.1485 } }, "supported_sizes": [ "1k", "2k", "4k" ], "benchmarks": { "elo": 1254, "rank": 3, "ci95": "-11/11", "appearances": 4833, "categories": { "style_general_photorealistic": { "elo": 1291, "ci95": "-18/18", "appearances": 1951 }, "style_cartoon_illustration": { "elo": 1260, "ci95": "-38/38", "appearances": 418 }, "subject_people:_portraits": { "elo": 1280, "ci95": "-32/32", "appearances": 607 }, "style_anime": { "elo": 1206, "ci95": "-37/37", "appearances": 409 }, "subject_futuristic_sci-fi": { "elo": 1237, "ci95": "-30/30", "appearances": 638 }, "subject_physical_spaces": { "elo": 1246, "ci95": "-30/30", "appearances": 672 }, "subject_text_typography": { "elo": 1163, "ci95": "-51/51", "appearances": 207 }, "subject_nature_landscapes": { "elo": 1253, "ci95": "-33/33", "appearances": 527 }, "style_vintage_retro": { "elo": 1274, "ci95": "-56/56", "appearances": 201 }, "subject_fantasy_mythical": { "elo": 1275, "ci95": "-27/27", "appearances": 764 }, "subject_ui/ux_design": { "elo": 1187, "ci95": "-72/72", "appearances": 126 }, "subject_commercial": { "elo": 1288, "ci95": "-51/51", "appearances": 248 }, "style_traditional_art": { "elo": 1326, "ci95": "-42/42", "appearances": 341 }, "style_graphic_design_digital_rendering": { "elo": 1204, "ci95": "-20/20", "appearances": 1349 }, "subject_people:_groups_activities": { "elo": 1322, "ci95": "-34/34", "appearances": 564 } } } }, { "id": "minimax-image-01", "name": "MiniMax: MiniMax Image-01", "short_name": "MiniMax Image-01", "description": "MiniMax Image-01 generates high-quality images from text.", "created": 1768953600, "owned_by": "minimax", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/images/generations", "/v1/images/edits" ], "pricing": { "unit": "per image", "resolutions": { "1024x1024": 0.0055, "1280x720": 0.0055, "1152x864": 0.0055, "1248x832": 0.0055, "832x1248": 0.0055, "864x1152": 0.0055, "720x1280": 0.0055, "1344x576": 0.0055, "auto": 0.0055 } }, "supported_sizes": [ "1024x1024", "1280x720", "1152x864", "1248x832", "832x1248", "864x1152", "720x1280", "1344x576" ], "benchmarks": null }, { "id": "flux-2-turbo", "name": "Black Forest Labs: FLUX.2 [turbo]", "short_name": "FLUX.2 [turbo]", "description": "FLUX.2 [turbo] is a speed-optimized image model for real-time workflows with strong prompt adherence and clean typography.", "created": 1766361600, "owned_by": "blackforestlabs", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/images/generations", "/v1/images/edits" ], "pricing": { "unit": "per image", "resolutions": { "1024x1024": 0.0187, "1280x720": 0.0187, "720x1280": 0.0187, "1536x1024": 0.0187, "1024x1536": 0.0187, "auto": 0.0187 } }, "supported_sizes": [ "1024x1024", "1280x720", "720x1280", "1536x1024", "1024x1536" ], "benchmarks": null }, { "id": "flux-2-flash", "name": "Black Forest Labs: FLUX.2 [flash]", "short_name": "FLUX.2 [flash]", "description": "FLUX.2 [flash] is a fast text-to-image model built for low-latency, high-volume generation with realistic renders and crisp on-image text. A strong default for rapid iteration, batch pipelines, and production creatives.", "created": 1766361600, "owned_by": "blackforestlabs", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/images/generations", "/v1/images/edits" ], "pricing": { "unit": "per image", "resolutions": { "1024x1024": 0.01496, "1280x720": 0.01496, "720x1280": 0.01496, "1536x1024": 0.01496, "1024x1536": 0.01496, "auto": 0.01496 } }, "supported_sizes": [ "1024x1024", "1280x720", "720x1280", "1536x1024", "1024x1536" ], "benchmarks": null }, { "id": "flux-2-dev-lora", "name": "Black Forest Labs: FLUX.2 [dev] LoRA", "short_name": "FLUX.2 [dev] LoRA", "description": "FLUX.2 [dev] with LoRA support delivers fast, studio-quality text-to-image generation with enhanced realism, crisper text rendering, and personalized styles via custom LoRA adapters. Supports up to 4 LoRAs for brand-specific outputs.", "created": 1704067200, "owned_by": "blackforestlabs", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/images/generations", "/v1/images/edits" ], "pricing": { "unit": "per image", "resolutions": { "1024x1024": 0.03366, "1280x720": 0.03366, "720x1280": 0.03366, "1536x1024": 0.03366, "1024x1536": 0.03366, "auto": 0.03366 } }, "supported_sizes": [ "1024x1024", "1280x720", "720x1280", "1536x1024", "1024x1536" ], "benchmarks": null }, { "id": "flux-2-flex", "name": "Black Forest Labs: FLUX.2 [flex]", "short_name": "FLUX.2 [flex]", "description": "FLUX.2 [flex] from Black Forest Labs. Lightweight open-source generation that maintains professional quality at high speed, delivering efficiency without compromising output fidelity. Enhanced realism, crisper text generation, and native editing capabilities.", "created": 1704067200, "owned_by": "blackforestlabs", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/images/generations", "/v1/images/edits" ], "pricing": { "unit": "per image", "resolutions": { "1024x1024": 0.04488, "1280x720": 0.04488, "720x1280": 0.04488, "1536x1024": 0.04488, "1024x1536": 0.04488, "auto": 0.04488 } }, "supported_sizes": [ "1024x1024", "1280x720", "720x1280", "1536x1024", "1024x1536" ], "benchmarks": null }, { "id": "flux-2-pro", "name": "Black Forest Labs: FLUX.2 [pro]", "short_name": "FLUX.2 [pro]", "description": "FLUX.2 [pro] from Black Forest Labs delivers enhanced realism, crisper text generation, and native editing capabilities. A lightweight open-source model optimized for rapid prototyping, high-volume generation, and custom LoRA fine-tuning workflows.", "created": 1704067200, "owned_by": "blackforestlabs", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/images/generations", "/v1/images/edits" ], "pricing": { "unit": "per image", "resolutions": { "1024x1024": 0.0561, "1280x720": 0.0561, "720x1280": 0.0561, "1536x1024": 0.0561, "1024x1536": 0.0561, "auto": 0.0561 } }, "supported_sizes": [ "1024x1024", "1280x720", "720x1280", "1536x1024", "1024x1536" ], "benchmarks": null }, { "id": "flux-2-max", "name": "Black Forest Labs: FLUX.2 [max]", "short_name": "FLUX.2 [max]", "description": "FLUX.2 [max] from Black Forest Labs delivers production-grade text-to-image generation with enhanced realism, sharper text rendering, and native editing for reliable, repeatable results. A flagship model tuned for professional-quality images without parameter hassle.", "created": 1765929600, "owned_by": "blackforestlabs", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/images/generations", "/v1/images/edits" ], "pricing": { "unit": "per image", "resolutions": { "1024x1024": 0.0561, "1280x720": 0.0561, "720x1280": 0.0561, "1536x1024": 0.0561, "1024x1536": 0.0561, "auto": 0.0561 } }, "supported_sizes": [ "1024x1024", "1280x720", "720x1280", "1536x1024", "1024x1536" ], "benchmarks": { "elo": 1200, "rank": 5, "ci95": "-12/12", "appearances": 3839, "categories": { "style_general_photorealistic": { "elo": 1217, "ci95": "-20/20", "appearances": 1372 }, "style_cartoon_illustration": { "elo": 1255, "ci95": "-46/46", "appearances": 298 }, "subject_people:_portraits": { "elo": 1187, "ci95": "-38/38", "appearances": 390 }, "style_anime": { "elo": 1219, "ci95": "-47/47", "appearances": 269 }, "subject_futuristic_sci-fi": { "elo": 1166, "ci95": "-32/32", "appearances": 499 }, "subject_physical_spaces": { "elo": 1161, "ci95": "-32/32", "appearances": 512 }, "subject_text_typography": { "elo": 1191, "ci95": "-57/57", "appearances": 178 }, "subject_nature_landscapes": { "elo": 1190, "ci95": "-39/39", "appearances": 342 }, "style_vintage_retro": { "elo": 1254, "ci95": "-57/57", "appearances": 196 }, "subject_fantasy_mythical": { "elo": 1241, "ci95": "-33/33", "appearances": 515 }, "subject_ui/ux_design": { "elo": 1216, "ci95": "-68/68", "appearances": 162 }, "subject_commercial": { "elo": 1223, "ci95": "-46/46", "appearances": 277 }, "style_traditional_art": { "elo": 1202, "ci95": "-46/46", "appearances": 240 }, "style_graphic_design_digital_rendering": { "elo": 1171, "ci95": "-21/21", "appearances": 1225 }, "subject_people:_groups_activities": { "elo": 1294, "ci95": "-43/43", "appearances": 344 } } } }, { "id": "z-image-base", "name": "Z.ai: Z Image Base", "short_name": "Z Image Base", "description": "Z Image Base is a 6B-parameter model that supports text-to-image.", "created": 1769558400, "owned_by": "z-ai", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/images/generations", "/v1/images/edits" ], "pricing": { "unit": "per image", "resolutions": { "256x256": 0.0187, "512x512": 0.0187, "768x768": 0.0187, "1024x1024": 0.0187, "1024x768": 0.0187, "768x1024": 0.0187, "1024x576": 0.0187, "576x1024": 0.0187, "auto": 0.0187 } }, "supported_sizes": [ "1024x1024", "1024x768", "768x1024", "1024x576", "576x1024", "768x768", "512x512", "256x256" ], "benchmarks": null }, { "id": "z-image-turbo", "name": "Z.ai: Z Image Turbo", "short_name": "Z Image Turbo", "description": "Z Image Turbo is a fast, high-quality image generation model optimized for speed. Generate detailed images with cinematic quality, film grain effects, and artistic styles.", "created": 1764201600, "owned_by": "z-ai", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/images/generations", "/v1/images/edits" ], "pricing": { "unit": "per image", "resolutions": { "256x256": 0.01309, "512x512": 0.01309, "768x768": 0.01309, "1024x1024": 0.01309, "1280x720": 0.01309, "720x1280": 0.01309, "1536x1024": 0.01309, "1024x1536": 0.01309, "1536x1536": 0.01309, "auto": 0.01309 } }, "supported_sizes": [ "256x256", "512x512", "768x768", "1024x1024", "1280x720", "720x1280", "1536x1024", "1024x1536", "1536x1536" ], "benchmarks": null }, { "id": "z-image-turbo-lora", "name": "Z.ai: Z Image Turbo LoRA", "short_name": "Z Image Turbo LoRA", "description": "Z Image Turbo LoRA supports up to 3 LoRAs for custom styles, characters, or brand identity.", "created": 1764201600, "owned_by": "z-ai", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/images/generations", "/v1/images/edits" ], "pricing": { "unit": "per image", "resolutions": { "256x256": 0.0187, "512x512": 0.0187, "768x768": 0.0187, "1024x1024": 0.0187, "1280x720": 0.0187, "720x1280": 0.0187, "1536x1024": 0.0187, "1024x1536": 0.0187, "1536x1536": 0.0187, "auto": 0.0187 } }, "supported_sizes": [ "256x256", "512x512", "768x768", "1024x1024", "1280x720", "720x1280", "1536x1024", "1024x1536", "1536x1536" ], "benchmarks": null }, { "id": "vidu-q2", "name": "Vidu: Vidu Q2", "short_name": "Vidu Q2", "description": "Vidu Q2 is a high-end text-to-image model with cinematic lighting and clean composition. Supports up to 4K resolution and flexible aspect ratios.", "created": 1764633600, "owned_by": "vidu", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/images/generations", "/v1/images/edits" ], "pricing": { "unit": "per image", "resolutions": { "1080p": 0.033, "2K": 0.044, "4K": 0.055, "auto": 0.033 } }, "supported_sizes": [ "1080p", "2K", "4K" ], "benchmarks": { "elo": 1113, "rank": 35, "ci95": "-9/9", "appearances": 5402, "categories": { "style_general_photorealistic": { "elo": 1132, "ci95": "-15/15", "appearances": 2186 }, "style_cartoon_illustration": { "elo": 1179, "ci95": "-33/33", "appearances": 492 }, "subject_people:_portraits": { "elo": 1155, "ci95": "-28/28", "appearances": 658 }, "style_anime": { "elo": 1145, "ci95": "-33/33", "appearances": 481 }, "subject_futuristic_sci-fi": { "elo": 1070, "ci95": "-25/25", "appearances": 751 }, "subject_physical_spaces": { "elo": 1129, "ci95": "-26/26", "appearances": 722 }, "subject_text_typography": { "elo": 1056, "ci95": "-54/54", "appearances": 181 }, "subject_nature_landscapes": { "elo": 1089, "ci95": "-29/29", "appearances": 566 }, "style_vintage_retro": { "elo": 1057, "ci95": "-48/48", "appearances": 226 }, "subject_fantasy_mythical": { "elo": 1186, "ci95": "-24/24", "appearances": 877 }, "subject_ui/ux_design": { "elo": 964, "ci95": "-58/58", "appearances": 165 }, "subject_commercial": { "elo": 1059, "ci95": "-43/43", "appearances": 271 }, "style_traditional_art": { "elo": 1199, "ci95": "-36/36", "appearances": 369 }, "style_graphic_design_digital_rendering": { "elo": 1057, "ci95": "-18/18", "appearances": 1456 }, "subject_people:_groups_activities": { "elo": 1216, "ci95": "-30/30", "appearances": 593 } } } }, { "id": "grok-2-image", "name": "xAI: Grok 2 Image", "short_name": "Grok 2 Image", "description": "Grok 2 Image is xAI's flagship image generation model that turns text prompts into sharp, photorealistic visuals. Optimized for marketing creatives, social posts, product visuals, and concept art with strong prompt following and flexible visual styles.", "created": 1765238400, "owned_by": "xai", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/images/generations", "/v1/images/edits" ], "pricing": { "unit": "per image", "resolutions": { "default": 0.1309, "auto": 0.1309 } }, "supported_sizes": [ "default" ], "benchmarks": null }, { "id": "seedream-v3", "name": "ByteDance: SeeDream 3.0", "short_name": "SeeDream 3.0", "description": "Generate images using Bytedance's SeedDream 3.0 model. Native 2K high resolution output with exceptional text layout, accurate text generation, photorealistic portraits, and fast generation.", "created": 1704067200, "owned_by": "bytedance", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/images/generations", "/v1/images/edits" ], "pricing": { "unit": "per image", "resolutions": { "1024x1024": 0.044, "1344x768": 0.044, "768x1344": 0.044, "1152x896": 0.044, "896x1152": 0.044, "1536x672": 0.044, "672x1536": 0.044, "2016x672": 0.044, "auto": 0.044 } }, "supported_sizes": [ "1024x1024", "1152x896", "896x1152", "1344x768", "768x1344", "1536x672", "672x1536", "2016x672" ], "benchmarks": { "elo": 1150, "rank": 17, "ci95": "-10/10", "appearances": 4287, "categories": { "style_general_photorealistic": { "elo": 1211, "ci95": "-17/17", "appearances": 1748 }, "style_cartoon_illustration": { "elo": 1116, "ci95": "-36/36", "appearances": 373 }, "subject_people:_portraits": { "elo": 1166, "ci95": "-32/32", "appearances": 484 }, "style_anime": { "elo": 1132, "ci95": "-41/41", "appearances": 309 }, "subject_futuristic_sci-fi": { "elo": 1154, "ci95": "-30/30", "appearances": 551 }, "subject_physical_spaces": { "elo": 1152, "ci95": "-29/29", "appearances": 592 }, "subject_text_typography": { "elo": 1043, "ci95": "-59/59", "appearances": 144 }, "subject_nature_landscapes": { "elo": 1208, "ci95": "-34/34", "appearances": 434 }, "style_vintage_retro": { "elo": 1068, "ci95": "-52/52", "appearances": 187 }, "subject_fantasy_mythical": { "elo": 1206, "ci95": "-29/29", "appearances": 624 }, "subject_commercial": { "elo": 1119, "ci95": "-45/45", "appearances": 246 }, "style_traditional_art": { "elo": 1248, "ci95": "-44/44", "appearances": 260 }, "style_graphic_design_digital_rendering": { "elo": 1075, "ci95": "-20/20", "appearances": 1192 }, "subject_people:_groups_activities": { "elo": 1211, "ci95": "-34/34", "appearances": 440 } } } }, { "id": "qwen-image", "name": "Qwen: Qwen Image", "short_name": "Qwen Image", "description": "Latest release (v2509, 25-09). Qwen-Image is an image generation foundation model that excels at complex text rendering.", "created": 1754524800, "owned_by": "qwen", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/images/generations", "/v1/images/edits" ], "pricing": { "unit": "per image", "resolutions": { "1024x1024": 0.0275, "1024x1536": 0.0275, "1536x1024": 0.0275, "1536x1536": 0.0275, "768x1024": 0.0275, "1024x768": 0.0275, "512x512": 0.0275, "256x256": 0.0275, "auto": 0.0275 } }, "supported_sizes": [ "1024x1024", "1024x1536", "1536x1024", "1536x1536", "768x1024", "1024x768", "512x512", "256x256" ], "benchmarks": { "elo": 1059, "rank": 63, "ci95": "-11/11", "appearances": 3693, "categories": { "style_general_photorealistic": { "elo": 1054, "ci95": "-18/18", "appearances": 1560 }, "style_cartoon_illustration": { "elo": 1132, "ci95": "-40/40", "appearances": 312 }, "subject_people:_portraits": { "elo": 1016, "ci95": "-36/36", "appearances": 392 }, "style_anime": { "elo": 1159, "ci95": "-42/42", "appearances": 285 }, "subject_futuristic_sci-fi": { "elo": 1081, "ci95": "-33/33", "appearances": 442 }, "subject_physical_spaces": { "elo": 1041, "ci95": "-31/31", "appearances": 487 }, "subject_nature_landscapes": { "elo": 1062, "ci95": "-35/35", "appearances": 400 }, "style_vintage_retro": { "elo": 1060, "ci95": "-60/60", "appearances": 144 }, "subject_fantasy_mythical": { "elo": 1138, "ci95": "-29/29", "appearances": 577 }, "subject_commercial": { "elo": 1005, "ci95": "-49/49", "appearances": 208 }, "style_traditional_art": { "elo": 1118, "ci95": "-46/46", "appearances": 228 }, "style_graphic_design_digital_rendering": { "elo": 1009, "ci95": "-22/22", "appearances": 946 }, "subject_people:_groups_activities": { "elo": 1134, "ci95": "-35/35", "appearances": 402 } } } }, { "id": "imagen-4", "name": "Google: Imagen 4", "short_name": "Imagen 4", "description": "Google's Imagen 4 model. Generates high-quality images with excellent detail and composition. Supports diverse art styles from photorealism to animation.", "created": 1747785600, "owned_by": "google", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/images/generations", "/v1/images/edits" ], "pricing": { "unit": "per image", "resolutions": { "1:1": 0.04, "16:9": 0.04, "9:16": 0.04, "3:4": 0.04, "4:3": 0.04, "auto": 0.04 } }, "supported_sizes": [ "1:1", "16:9", "9:16", "3:4", "4:3" ], "benchmarks": { "elo": 1065, "rank": 61, "ci95": "-12/12", "appearances": 3353, "categories": { "style_general_photorealistic": { "elo": 1106, "ci95": "-19/19", "appearances": 1341 }, "style_cartoon_illustration": { "elo": 1040, "ci95": "-44/44", "appearances": 262 }, "subject_people:_portraits": { "elo": 1072, "ci95": "-34/34", "appearances": 415 }, "style_anime": { "elo": 1066, "ci95": "-44/44", "appearances": 251 }, "subject_futuristic_sci-fi": { "elo": 1016, "ci95": "-32/32", "appearances": 487 }, "subject_physical_spaces": { "elo": 1073, "ci95": "-33/33", "appearances": 440 }, "subject_text_typography": { "elo": 1125, "ci95": "-64/64", "appearances": 131 }, "subject_nature_landscapes": { "elo": 1112, "ci95": "-39/39", "appearances": 320 }, "style_vintage_retro": { "elo": 1040, "ci95": "-53/53", "appearances": 177 }, "subject_fantasy_mythical": { "elo": 1082, "ci95": "-30/30", "appearances": 526 }, "subject_commercial": { "elo": 1066, "ci95": "-49/49", "appearances": 207 }, "style_traditional_art": { "elo": 1134, "ci95": "-47/47", "appearances": 217 }, "style_graphic_design_digital_rendering": { "elo": 1010, "ci95": "-22/22", "appearances": 963 }, "subject_people:_groups_activities": { "elo": 1112, "ci95": "-40/40", "appearances": 306 } } }, "discount": { "percentage": 5, "note": "A 5% discount applies to input and output token pricing only." } }, { "id": "hidream", "name": "Hidream", "short_name": "Hidream", "description": "Hidream I1 Full, the latest and greatest image generation model from Hidream.", "created": 1704067200, "owned_by": "hidream", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/images/generations", "/v1/images/edits" ], "pricing": { "unit": "per image", "resolutions": { "1024x1024": 0.01683, "768x1360": 0.01683, "1360x768": 0.01683, "880x1168": 0.01683, "1168x880": 0.01683, "1248x832": 0.01683, "832x1248": 0.01683, "1024x768": 0.01683, "768x1024": 0.01683, "1024x576": 0.01683, "576x1024": 0.01683, "1408x1024": 0.01683, "1024x1408": 0.01683, "512x512": 0.01683, "2048x2048": 0.01683, "1920x1088": 0.01683, "1088x1920": 0.01683, "auto": 0.01683 } }, "supported_sizes": [ "1024x1024", "768x1360", "1360x768", "880x1168", "1168x880", "1248x832", "832x1248", "1024x768", "768x1024", "1024x576", "576x1024", "1408x1024", "1024x1408", "512x512", "2048x2048", "1920x1088", "1088x1920" ], "benchmarks": null }, { "id": "flux-1-crea-dev", "name": "Black Forest Labs: Flux 1 Krea Dev", "short_name": "Flux 1 Krea Dev", "description": "Same as Flux Dev in every way except it's Flux Krea. High-quality image generation with excellent prompt adherence.", "created": 1704067200, "owned_by": "blackforestlabs", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/images/generations", "/v1/images/edits" ], "pricing": { "unit": "per image", "resolutions": { "auto": 0.01309, "1024x1024": 0.01309, "1920x1088": 0.01309, "1088x1920": 0.01309, "768x1024": 0.01309, "1024x768": 0.01309, "1408x1024": 0.01309, "1024x1408": 0.01309, "512x512": 0.01309, "2048x2048": 0.01309 } }, "benchmarks": null, "supported_sizes": [ "1024x1024", "1920x1088", "1088x1920", "768x1024", "1024x768", "1408x1024", "1024x1408", "512x512", "2048x2048" ] }, { "id": "ideogram-v3-quality", "name": "Ideogram: Ideogram V3 Quality", "short_name": "Ideogram V3 Quality", "description": "Ideogram V3 (Quality) offers the highest fidelity images of the Ideogram series and poster-grade text rendering capabilities.", "created": 1704067200, "owned_by": "ideogram", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/images/generations", "/v1/images/edits" ], "pricing": { "unit": "per image", "resolutions": { "1024x1024": 0.099, "768x1024": 0.099, "1024x768": 0.099, "640x1344": 0.099, "1344x640": 0.099, "512x1536": 0.099, "1536x512": 0.099, "auto": 0.099 } }, "supported_sizes": [ "1024x1024", "768x1024", "1024x768", "640x1344", "1344x640", "512x1536", "1536x512" ], "benchmarks": null }, { "id": "recraft-v3", "name": "Recraft: Recraft V3", "short_name": "Recraft V3", "description": "Recraft V3 is a state-of-the-art image generation model that is known for its high quality and prompt adherence.", "created": 1704067200, "owned_by": "recraft", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/images/generations", "/v1/images/edits" ], "pricing": { "unit": "per image", "resolutions": { "1024x1024": 0.044, "1024x768": 0.044, "1024x576": 0.044, "768x1024": 0.044, "576x1024": 0.044, "auto": 0.044 } }, "supported_sizes": [ "1024x1024", "1024x768", "1024x576", "768x1024", "576x1024" ], "benchmarks": { "elo": 1073, "rank": 51, "ci95": "-9/9", "appearances": 6382, "categories": { "style_general_photorealistic": { "elo": 1123, "ci95": "-13/13", "appearances": 2811 }, "style_cartoon_illustration": { "elo": 987, "ci95": "-32/32", "appearances": 512 }, "subject_people:_portraits": { "elo": 1063, "ci95": "-31/31", "appearances": 517 }, "style_anime": { "elo": 949, "ci95": "-36/36", "appearances": 419 }, "subject_futuristic_sci-fi": { "elo": 1059, "ci95": "-24/24", "appearances": 835 }, "subject_physical_spaces": { "elo": 1101, "ci95": "-22/22", "appearances": 1087 }, "subject_text_typography": { "elo": 1078, "ci95": "-44/44", "appearances": 282 }, "subject_nature_landscapes": { "elo": 1137, "ci95": "-26/26", "appearances": 738 }, "style_vintage_retro": { "elo": 1088, "ci95": "-46/46", "appearances": 256 }, "subject_fantasy_mythical": { "elo": 1038, "ci95": "-22/22", "appearances": 1018 }, "subject_ui/ux_design": { "elo": 969, "ci95": "-40/40", "appearances": 346 }, "subject_commercial": { "elo": 1074, "ci95": "-37/37", "appearances": 379 }, "style_traditional_art": { "elo": 1121, "ci95": "-37/37", "appearances": 367 }, "style_graphic_design_digital_rendering": { "elo": 1050, "ci95": "-16/16", "appearances": 1845 }, "subject_people:_groups_activities": { "elo": 1104, "ci95": "-29/29", "appearances": 625 } } } }, { "id": "recraft-v4", "name": "Recraft: Recraft V4", "short_name": "Recraft V4", "description": "Recraft V4 is a text-to-image model designed for professional design and marketing use cases.", "created": 1704067200, "owned_by": "recraft", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/images/generations", "/v1/images/edits" ], "pricing": { "unit": "per image", "resolutions": { "1024x1024": 0.044, "1024x768": 0.044, "1024x576": 0.044, "768x1024": 0.044, "576x1024": 0.044, "auto": 0.044 } }, "supported_sizes": [ "1024x1024", "1024x768", "1024x576", "768x1024", "576x1024" ], "benchmarks": { "elo": 1121, "rank": 31, "ci95": "-9/9", "appearances": 5132, "categories": { "style_general_photorealistic": { "elo": 1154, "ci95": "-16/16", "appearances": 1764 }, "style_cartoon_illustration": { "elo": 1097, "ci95": "-36/36", "appearances": 374 }, "subject_people:_portraits": { "elo": 1109, "ci95": "-30/30", "appearances": 549 }, "style_anime": { "elo": 1053, "ci95": "-37/37", "appearances": 360 }, "subject_futuristic_sci-fi": { "elo": 1123, "ci95": "-28/28", "appearances": 614 }, "subject_physical_spaces": { "elo": 1111, "ci95": "-27/27", "appearances": 634 }, "subject_text_typography": { "elo": 1099, "ci95": "-45/45", "appearances": 245 }, "subject_nature_landscapes": { "elo": 1117, "ci95": "-33/33", "appearances": 432 }, "style_vintage_retro": { "elo": 1116, "ci95": "-45/45", "appearances": 241 }, "subject_fantasy_mythical": { "elo": 1174, "ci95": "-26/26", "appearances": 686 }, "subject_ui/ux_design": { "elo": 1111, "ci95": "-55/55", "appearances": 173 }, "subject_commercial": { "elo": 1142, "ci95": "-40/40", "appearances": 297 }, "style_traditional_art": { "elo": 1237, "ci95": "-39/39", "appearances": 319 }, "style_graphic_design_digital_rendering": { "elo": 1093, "ci95": "-18/18", "appearances": 1380 }, "subject_people:_groups_activities": { "elo": 1136, "ci95": "-31/31", "appearances": 513 } } } }, { "id": "recraft-v4-pro", "name": "Recraft: Recraft V4 Pro", "short_name": "Recraft V4 Pro", "description": "Recraft V4 Pro is a professional text-to-image model for premium design and marketing output.", "created": 1704067200, "owned_by": "recraft", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/images/generations", "/v1/images/edits" ], "pricing": { "unit": "per image", "resolutions": { "1024x1024": 0.275, "1024x768": 0.275, "1024x576": 0.275, "768x1024": 0.275, "576x1024": 0.275, "auto": 0.275 } }, "supported_sizes": [ "1024x1024", "1024x768", "1024x576", "768x1024", "576x1024" ], "benchmarks": { "elo": 1132, "rank": 26, "ci95": "-9/9", "appearances": 5042, "categories": { "style_general_photorealistic": { "elo": 1173, "ci95": "-16/16", "appearances": 1774 }, "style_cartoon_illustration": { "elo": 1103, "ci95": "-35/35", "appearances": 407 }, "subject_people:_portraits": { "elo": 1126, "ci95": "-31/31", "appearances": 510 }, "style_anime": { "elo": 1087, "ci95": "-36/36", "appearances": 381 }, "subject_futuristic_sci-fi": { "elo": 1095, "ci95": "-28/28", "appearances": 618 }, "subject_physical_spaces": { "elo": 1108, "ci95": "-27/27", "appearances": 633 }, "subject_text_typography": { "elo": 1089, "ci95": "-48/48", "appearances": 213 }, "subject_nature_landscapes": { "elo": 1129, "ci95": "-33/33", "appearances": 421 }, "style_vintage_retro": { "elo": 1098, "ci95": "-46/46", "appearances": 238 }, "subject_fantasy_mythical": { "elo": 1162, "ci95": "-27/27", "appearances": 664 }, "subject_ui/ux_design": { "elo": 1100, "ci95": "-54/54", "appearances": 178 }, "subject_commercial": { "elo": 1139, "ci95": "-39/39", "appearances": 320 }, "style_traditional_art": { "elo": 1234, "ci95": "-41/41", "appearances": 291 }, "style_graphic_design_digital_rendering": { "elo": 1079, "ci95": "-19/19", "appearances": 1273 }, "subject_people:_groups_activities": { "elo": 1201, "ci95": "-32/32", "appearances": 498 } } } }, { "id": "sdxl-arlimix-v1", "name": "Stability: SDXL ArliMix V1", "short_name": "SDXL ArliMix V1", "description": "Image generation using SDXL ArliMix V1 via Arli AI. Cartoon-like generation.", "created": 1704067200, "owned_by": "stability", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/images/generations", "/v1/images/edits" ], "pricing": { "unit": "per image", "resolutions": { "512x512": 0.00935, "768x768": 0.00935, "1024x1024": 0.00935, "1408x1408": 0.00935, "576x1024": 0.00935, "1024x576": 0.00935, "768x1024": 0.00935, "1024x768": 0.00935, "auto": 0.00935 } }, "benchmarks": null, "supported_sizes": [ "512x512", "768x768", "1024x1024", "1408x1408", "576x1024", "1024x576", "768x1024", "1024x768" ] }, { "id": "grok-imagine-image", "name": "xAI: Grok Imagine Image", "short_name": "Grok Imagine Image", "description": "Generate highly aesthetic images with xAI's Grok Imagine Image model. ", "created": 1769644800, "owned_by": "xai", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/images/generations", "/v1/images/edits" ], "pricing": { "unit": "per image", "resolutions": { "2:1": 0.022, "20:9": 0.022, "19.5:9": 0.022, "16:9": 0.022, "4:3": 0.022, "3:2": 0.022, "1:1": 0.022, "2:3": 0.022, "3:4": 0.022, "9:16": 0.022, "9:19.5": 0.022, "9:20": 0.022, "1:2": 0.022, "auto": 0.022 } }, "supported_sizes": [ "2:1", "20:9", "19.5:9", "16:9", "4:3", "3:2", "1:1", "2:3", "3:4", "9:16", "9:19.5", "9:20", "1:2" ], "benchmarks": { "elo": 1171, "rank": 11, "ci95": "-8/8", "appearances": 6717, "categories": { "style_general_photorealistic": { "elo": 1172, "ci95": "-14/14", "appearances": 2352 }, "style_cartoon_illustration": { "elo": 1180, "ci95": "-30/30", "appearances": 541 }, "subject_people:_portraits": { "elo": 1172, "ci95": "-26/26", "appearances": 716 }, "style_anime": { "elo": 1150, "ci95": "-31/31", "appearances": 501 }, "subject_futuristic_sci-fi": { "elo": 1190, "ci95": "-24/24", "appearances": 872 }, "subject_physical_spaces": { "elo": 1148, "ci95": "-24/24", "appearances": 830 }, "subject_text_typography": { "elo": 1229, "ci95": "-42/42", "appearances": 317 }, "subject_nature_landscapes": { "elo": 1131, "ci95": "-27/27", "appearances": 629 }, "style_vintage_retro": { "elo": 1172, "ci95": "-40/40", "appearances": 316 }, "subject_fantasy_mythical": { "elo": 1195, "ci95": "-24/24", "appearances": 875 }, "subject_ui/ux_design": { "elo": 1164, "ci95": "-48/48", "appearances": 238 }, "subject_commercial": { "elo": 1182, "ci95": "-35/35", "appearances": 411 }, "style_traditional_art": { "elo": 1229, "ci95": "-35/35", "appearances": 396 }, "style_graphic_design_digital_rendering": { "elo": 1172, "ci95": "-16/16", "appearances": 1918 }, "subject_people:_groups_activities": { "elo": 1200, "ci95": "-28/28", "appearances": 639 } } } }, { "id": "flux-1.1-pro", "name": "Black Forest Labs: Flux Pro V1.1 Ultra", "short_name": "Flux Pro V1.1 Ultra", "description": "Excellent image quality, prompt adherence, and output diversity.", "created": 1704067200, "owned_by": "blackforestlabs", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/images/generations", "/v1/images/edits" ], "pricing": { "unit": "per image", "resolutions": { "2048x2048": 0.066, "2752x1536": 0.066, "1536x2752": 0.066, "2304x1792": 0.066, "1792x2304": 0.066, "3136x1344": 0.066, "1344x3136": 0.066, "auto": 0.066 } }, "benchmarks": null, "supported_sizes": [ "2048x2048", "3136x1344", "2752x1536", "2304x1792", "1792x2304", "1536x2752", "1344x3136" ] }, { "id": "chroma", "name": "WaveSpeed: Chroma", "short_name": "Chroma", "description": "Uncensored text-to-image model via Wavespeed.", "created": 1754956800, "owned_by": "wavespeed", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/images/generations", "/v1/images/edits" ], "pricing": { "unit": "per image", "resolutions": { "1024x1024": 0.02805, "512x512": 0.02805, "768x1024": 0.02805, "576x1024": 0.02805, "1024x768": 0.02805, "1024x576": 0.02805, "1536x1536": 0.02805, "1536x1024": 0.02805, "1024x1536": 0.02805, "auto": 0.02805 } }, "supported_sizes": [ "1536x1536", "1536x1024", "1024x1536", "1024x1024", "512x512", "768x1024", "576x1024", "1024x768", "1024x576" ], "benchmarks": null }, { "id": "ideogram-v2", "name": "Ideogram: Ideogram V2", "short_name": "Ideogram V2", "description": "An excellent image model with state of the art inpainting, prompt comprehension and especially text rendering.", "created": 1704067200, "owned_by": "ideogram", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/images/generations", "/v1/images/edits" ], "pricing": { "unit": "per image", "resolutions": { "1024x1024": 0.1496, "768x1024": 0.1496, "1024x768": 0.1496, "640x1344": 0.1496, "1344x640": 0.1496, "512x1536": 0.1496, "1536x512": 0.1496, "auto": 0.1496 } }, "benchmarks": { "elo": 1057, "rank": 64, "ci95": "-8/8", "appearances": 8113, "categories": { "style_general_photorealistic": { "elo": 1093, "ci95": "-12/12", "appearances": 3595 }, "style_cartoon_illustration": { "elo": 1024, "ci95": "-31/31", "appearances": 553 }, "subject_people:_portraits": { "elo": 1059, "ci95": "-32/32", "appearances": 512 }, "style_anime": { "elo": 981, "ci95": "-32/32", "appearances": 513 }, "subject_futuristic_sci-fi": { "elo": 1040, "ci95": "-22/22", "appearances": 1074 }, "subject_physical_spaces": { "elo": 1056, "ci95": "-18/18", "appearances": 1523 }, "subject_text_typography": { "elo": 1069, "ci95": "-36/36", "appearances": 435 }, "subject_nature_landscapes": { "elo": 1129, "ci95": "-23/23", "appearances": 944 }, "style_vintage_retro": { "elo": 1086, "ci95": "-40/40", "appearances": 336 }, "subject_fantasy_mythical": { "elo": 1045, "ci95": "-21/21", "appearances": 1177 }, "subject_ui/ux_design": { "elo": 988, "ci95": "-33/33", "appearances": 518 }, "subject_commercial": { "elo": 1023, "ci95": "-33/33", "appearances": 487 }, "style_traditional_art": { "elo": 1148, "ci95": "-36/36", "appearances": 382 }, "style_graphic_design_digital_rendering": { "elo": 1013, "ci95": "-14/14", "appearances": 2505 }, "subject_people:_groups_activities": { "elo": 1096, "ci95": "-27/27", "appearances": 725 } } }, "supported_sizes": [ "1024x1024", "768x1024", "1024x768", "640x1344", "1344x640", "512x1536", "1536x512" ] }, { "id": "flux-lightning", "name": "Black Forest Labs: Flux Lightning", "short_name": "Flux Lightning", "description": "Juggernaut Lightning by FAL, for the fastest text-to-image generation with high-quality results.", "created": 1704067200, "owned_by": "blackforestlabs", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/images/generations", "/v1/images/edits" ], "pricing": { "unit": "per image", "resolutions": { "1024x1024": 0.01122, "1024x768": 0.00935, "1024x576": 0.00748, "768x1024": 0.00935, "576x1024": 0.00748, "auto": 0.00748 } }, "supported_sizes": [ "1024x1024", "1024x768", "1024x576", "768x1024", "576x1024" ], "benchmarks": null }, { "id": "stable-diffusion-v3.5-large", "name": "Stability: Stable Diffusion 3.5 Large", "short_name": "Stable Diffusion 3.5 Large", "description": "Stable Diffusion's newest model. Generates a wide variety of images reflecting different styles without complex prompting.", "created": 1704067200, "owned_by": "stability", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/images/generations", "/v1/images/edits" ], "pricing": { "unit": "per image", "resolutions": { "1792x1024": 0.1271, "1024x1792": 0.1271, "1024x768": 0.1271, "768x1024": 0.1271, "1080x1350": 0.1271, "auto": 0.1271 } }, "benchmarks": { "elo": 1028, "rank": 78, "ci95": "-9/9", "appearances": 5550, "categories": { "style_general_photorealistic": { "elo": 1046, "ci95": "-14/14", "appearances": 2479 }, "style_cartoon_illustration": { "elo": 970, "ci95": "-36/36", "appearances": 414 }, "subject_people:_portraits": { "elo": 984, "ci95": "-36/36", "appearances": 404 }, "style_anime": { "elo": 1028, "ci95": "-37/37", "appearances": 371 }, "subject_futuristic_sci-fi": { "elo": 1049, "ci95": "-26/26", "appearances": 731 }, "subject_physical_spaces": { "elo": 1044, "ci95": "-22/22", "appearances": 1014 }, "subject_text_typography": { "elo": 969, "ci95": "-44/44", "appearances": 263 }, "subject_nature_landscapes": { "elo": 1086, "ci95": "-27/27", "appearances": 683 }, "style_vintage_retro": { "elo": 1010, "ci95": "-50/50", "appearances": 214 }, "subject_fantasy_mythical": { "elo": 1056, "ci95": "-25/25", "appearances": 821 }, "subject_ui/ux_design": { "elo": 947, "ci95": "-44/44", "appearances": 294 }, "subject_commercial": { "elo": 1003, "ci95": "-41/41", "appearances": 322 }, "style_traditional_art": { "elo": 1126, "ci95": "-42/42", "appearances": 284 }, "style_graphic_design_digital_rendering": { "elo": 1004, "ci95": "-17/17", "appearances": 1612 }, "subject_people:_groups_activities": { "elo": 1014, "ci95": "-33/33", "appearances": 496 } } }, "supported_sizes": [ "1792x1024", "1024x1792", "1024x768", "768x1024", "1080x1350" ] }, { "id": "ideogram-v2-turbo", "name": "Ideogram: Ideogram V2 Turbo", "short_name": "Ideogram V2 Turbo", "description": "A fast image model with state of the art inpainting, prompt comprehension and especially text rendering.", "created": 1704067200, "owned_by": "ideogram", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/images/generations", "/v1/images/edits" ], "pricing": { "unit": "per image", "resolutions": { "1024x1024": 0.0935, "768x1024": 0.0935, "1024x768": 0.0935, "640x1344": 0.0935, "1344x640": 0.0935, "512x1536": 0.0935, "1536x512": 0.0935, "auto": 0.0935 } }, "benchmarks": { "elo": 1043, "rank": 70, "ci95": "-8/8", "appearances": 7077, "categories": { "style_general_photorealistic": { "elo": 1059, "ci95": "-12/12", "appearances": 3133 }, "style_cartoon_illustration": { "elo": 1017, "ci95": "-32/32", "appearances": 522 }, "subject_people:_portraits": { "elo": 975, "ci95": "-34/34", "appearances": 446 }, "style_anime": { "elo": 961, "ci95": "-33/33", "appearances": 479 }, "subject_futuristic_sci-fi": { "elo": 1067, "ci95": "-22/22", "appearances": 1033 }, "subject_physical_spaces": { "elo": 1040, "ci95": "-20/20", "appearances": 1206 }, "subject_text_typography": { "elo": 1037, "ci95": "-39/39", "appearances": 362 }, "subject_nature_landscapes": { "elo": 1059, "ci95": "-24/24", "appearances": 846 }, "style_vintage_retro": { "elo": 1034, "ci95": "-47/47", "appearances": 238 }, "subject_fantasy_mythical": { "elo": 1019, "ci95": "-22/22", "appearances": 1063 }, "subject_ui/ux_design": { "elo": 1038, "ci95": "-38/38", "appearances": 410 }, "subject_commercial": { "elo": 1052, "ci95": "-36/36", "appearances": 421 }, "style_traditional_art": { "elo": 1061, "ci95": "-37/37", "appearances": 353 }, "style_graphic_design_digital_rendering": { "elo": 1048, "ci95": "-15/15", "appearances": 2151 }, "subject_people:_groups_activities": { "elo": 1083, "ci95": "-30/30", "appearances": 608 } } }, "supported_sizes": [ "1024x1024", "768x1024", "1024x768", "640x1344", "1344x640", "512x1536", "1536x512" ] }, { "id": "playground-v2.5", "name": "Playground: Playground V2.5", "short_name": "Playground V2.5", "description": "Playground V2.5 outperforms SDXL in many user tests. Suitable for a broad range of images.", "created": 1704067200, "owned_by": "playground", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/images/generations", "/v1/images/edits" ], "pricing": { "unit": "per image", "resolutions": { "1024x1024": 0.00935, "1024x768": 0.00935, "1024x576": 0.00935, "768x1024": 0.00935, "576x1024": 0.00935, "auto": 0.00935 } }, "benchmarks": { "elo": 970, "rank": 97, "ci95": "-6/6", "appearances": 12022, "categories": { "style_general_photorealistic": { "elo": 996, "ci95": "-9/9", "appearances": 5427 }, "style_cartoon_illustration": { "elo": 973, "ci95": "-25/25", "appearances": 859 }, "subject_people:_portraits": { "elo": 987, "ci95": "-31/31", "appearances": 557 }, "style_anime": { "elo": 964, "ci95": "-26/26", "appearances": 778 }, "subject_futuristic_sci-fi": { "elo": 991, "ci95": "-17/17", "appearances": 1662 }, "subject_physical_spaces": { "elo": 1011, "ci95": "-14/14", "appearances": 2383 }, "subject_text_typography": { "elo": 853, "ci95": "-33/33", "appearances": 484 }, "subject_nature_landscapes": { "elo": 1041, "ci95": "-18/18", "appearances": 1513 }, "style_vintage_retro": { "elo": 873, "ci95": "-34/34", "appearances": 464 }, "subject_fantasy_mythical": { "elo": 1005, "ci95": "-16/16", "appearances": 1917 }, "subject_ui/ux_design": { "elo": 827, "ci95": "-24/24", "appearances": 903 }, "subject_commercial": { "elo": 908, "ci95": "-27/27", "appearances": 715 }, "style_traditional_art": { "elo": 1089, "ci95": "-28/28", "appearances": 618 }, "style_graphic_design_digital_rendering": { "elo": 926, "ci95": "-11/11", "appearances": 3709 }, "subject_people:_groups_activities": { "elo": 938, "ci95": "-23/23", "appearances": 1039 } } }, "supported_sizes": [ "1024x1024", "1024x768", "1024x576", "768x1024", "576x1024" ] }, { "id": "ideogram-v3-turbo", "name": "Ideogram: Ideogram V3 Turbo", "short_name": "Ideogram V3 Turbo", "description": "Ideogram V3 (Turbo) generates images with lightning-fast speed and high text rendering accuracy.", "created": 1704067200, "owned_by": "ideogram", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/images/generations", "/v1/images/edits" ], "pricing": { "unit": "per image", "resolutions": { "1024x1024": 0.0561, "768x1024": 0.0561, "1024x768": 0.0561, "640x1344": 0.0561, "1344x640": 0.0561, "512x1536": 0.0561, "1536x512": 0.0561, "auto": 0.0561 } }, "supported_sizes": [ "1024x1024", "768x1024", "1024x768", "640x1344", "1344x640", "512x1536", "1536x512" ], "benchmarks": null }, { "id": "ideogram-v3-default", "name": "Ideogram: Ideogram V3", "short_name": "Ideogram V3", "description": "Ideogram V3 (Default) provides a good balance between speed and quality for text-to-image generation.", "created": 1704067200, "owned_by": "ideogram", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/images/generations", "/v1/images/edits" ], "pricing": { "unit": "per image", "resolutions": { "1024x1024": 0.099, "768x1024": 0.099, "1024x768": 0.099, "640x1344": 0.099, "1344x640": 0.099, "512x1536": 0.099, "1536x512": 0.099, "auto": 0.099 } }, "supported_sizes": [ "1024x1024", "768x1024", "1024x768", "640x1344", "1344x640", "512x1536", "1536x512" ], "benchmarks": null }, { "id": "hidream-i1-fast", "name": "HiDream: HiDream-I1 Fast", "short_name": "HiDream-I1 Fast", "description": "Fast image generation model. Optimized for speed while maintaining good quality.", "created": 1704067200, "owned_by": "hidream", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/images/generations", "/v1/images/edits" ], "pricing": { "unit": "per image", "resolutions": { "1024x1024": 0.00374, "1920x1080": 0.00561, "1080x1920": 0.00561, "768x1024": 0.00374, "1024x768": 0.00374, "auto": 0.00374 } }, "supported_sizes": [ "1024x1024", "1920x1080", "1080x1920", "768x1024", "1024x768" ], "benchmarks": { "elo": 1046, "rank": 68, "ci95": "-12/12", "appearances": 3136, "categories": { "style_general_photorealistic": { "elo": 1038, "ci95": "-20/20", "appearances": 1246 }, "style_cartoon_illustration": { "elo": 1071, "ci95": "-41/41", "appearances": 286 }, "subject_people:_portraits": { "elo": 1025, "ci95": "-36/36", "appearances": 385 }, "style_anime": { "elo": 1050, "ci95": "-48/48", "appearances": 218 }, "subject_futuristic_sci-fi": { "elo": 1041, "ci95": "-34/34", "appearances": 417 }, "subject_physical_spaces": { "elo": 1033, "ci95": "-34/34", "appearances": 428 }, "subject_nature_landscapes": { "elo": 997, "ci95": "-42/42", "appearances": 292 }, "subject_fantasy_mythical": { "elo": 1083, "ci95": "-32/32", "appearances": 467 }, "subject_commercial": { "elo": 1078, "ci95": "-53/53", "appearances": 175 }, "style_traditional_art": { "elo": 1117, "ci95": "-48/48", "appearances": 209 }, "style_graphic_design_digital_rendering": { "elo": 1026, "ci95": "-23/23", "appearances": 885 }, "subject_people:_groups_activities": { "elo": 1076, "ci95": "-41/41", "appearances": 301 } } } }, { "id": "juggernaut-pro-flux", "name": "Black Forest Labs: Juggernaut Pro Flux", "short_name": "Juggernaut Pro Flux", "description": "Premium quality image generation with Juggernaut Pro Flux. Delivers exceptional detail and artistic quality.", "created": 1704067200, "owned_by": "blackforestlabs", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/images/generations", "/v1/images/edits" ], "pricing": { "unit": "per image", "resolutions": { "1024x1024": 0.0187, "1920x1088": 0.0187, "1088x1920": 0.0187, "768x1024": 0.0187, "1024x768": 0.0187, "1408x1024": 0.0187, "1024x1408": 0.0187, "512x512": 0.0187, "2048x2048": 0.0187, "auto": 0.0187 } }, "supported_sizes": [ "1024x1024", "1920x1088", "1088x1920", "768x1024", "1024x768", "1408x1024", "1024x1408", "512x512", "2048x2048" ], "benchmarks": null }, { "id": "juggernaut-lightning-flux", "name": "Black Forest Labs: Juggernaut Lightning Flux", "short_name": "Juggernaut Lightning Flux", "description": "Ultra-fast image generation with Juggernaut Lightning Flux. Optimized for speed while maintaining good quality.", "created": 1704067200, "owned_by": "blackforestlabs", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/images/generations", "/v1/images/edits" ], "pricing": { "unit": "per image", "resolutions": { "1024x1024": 0.00935, "1920x1088": 0.01496, "1088x1920": 0.01496, "768x1024": 0.00935, "1024x768": 0.00935, "1408x1024": 0.01122, "1024x1408": 0.01122, "512x512": 0.00935, "2048x2048": 0.0187, "auto": 0.00935 } }, "supported_sizes": [ "1024x1024", "1920x1088", "1088x1920", "768x1024", "1024x768", "1408x1024", "1024x1408", "512x512", "2048x2048" ], "benchmarks": null }, { "id": "proteus", "name": "Proteus: Proteus", "short_name": "Proteus", "description": "Proteus model - high-quality image generation with excellent detail and artistic capabilities.", "created": 1704067200, "owned_by": "proteus", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/images/generations", "/v1/images/edits" ], "pricing": { "unit": "per image", "resolutions": { "1024x1024": 0.00374, "1024x768": 0.00374, "1024x576": 0.00374, "768x1024": 0.00374, "576x1024": 0.00374, "1408x1024": 0.00374, "1024x1408": 0.00374, "512x512": 0.00374, "2048x2048": 0.00374, "1920x1088": 0.00374, "1088x1920": 0.00374, "auto": 0.00374 } }, "supported_sizes": [ "1024x1024", "1920x1088", "1088x1920", "768x1024", "1024x768", "1408x1024", "1024x1408", "512x512", "2048x2048", "1024x576", "576x1024" ], "benchmarks": null }, { "id": "animagine-xl-4.0", "name": "Cagliostro Labs: Animagine XL 4.0", "short_name": "Animagine XL 4.0", "description": "Animagine XL 4.0 - High-quality anime generation model with NSFW capabilities.", "created": 1704067200, "owned_by": "cagliostrolab", "available": true, "hallucination_rate": 0, "endpoints": [ "/v1/images/generations", "/v1/images/edits" ], "pricing": { "unit": "per image", "resolutions": { "1024x1024": 0.00561, "1920x1088": 0.00561, "1088x1920": 0.00561, "768x1024": 0.00561, "1024x768": 0.00561, "1408x1024": 0.00561, "1024x1408": 0.00561, "512x512": 0.00374, "2048x2048": 0.00748, "768x768": 0.00468, "896x896": 0.00561, "1152x1152": 0.00561, "1280x1280": 0.00561, "1536x1536": 0.00655, "auto": 0.00374 } }, "benchmarks": null, "supported_sizes": [ "1024x1024", "1920x1088", "1088x1920", "768x1024", "1024x768", "1408x1024", "1024x1408", "512x512", "2048x2048", "768x768", "896x896", "1152x1152", "1280x1280", "1536x1536" ] } ] }