{"data":{"object":"list","data":[{"id":"claude-fable-5-1","family":"claude","family_label":"claude文本分组","endpoints":["anthropic","openai"],"context_length":1000000,"max_output_tokens":128000,"input_usd_per_million":5,"output_usd_per_million":25,"discount":0.5,"capabilities":["Reasoning","Tools","Files","Vision"],"description":"Claude model for demanding reasoning and long-horizon agentic work"},{"id":"claude-fable-5-2","family":"claude","family_label":"claude文本分组","endpoints":["anthropic","openai"],"context_length":1000000,"max_output_tokens":128000,"input_usd_per_million":4,"output_usd_per_million":20,"discount":0.5,"capabilities":["Reasoning","Tools","Files","Vision"],"description":"Claude model for demanding reasoning and long-horizon agentic work"},{"id":"claude-haiku-4-5-20251001","family":"claude","family_label":"claude文本分组","endpoints":["anthropic","openai"],"context_length":200000,"max_output_tokens":64000,"input_usd_per_million":0.5,"output_usd_per_million":2.5,"discount":0.5,"capabilities":["Reasoning","Tools","Files","Vision"],"description":"Fast Claude model for responsive assistance, classification, and lightweight agents"},{"id":"claude-haiku-5-5","family":"claude","family_label":"claude文本分组","endpoints":["anthropic","openai"],"context_length":1000000,"max_output_tokens":128000,"input_usd_per_million":0.05,"output_usd_per_million":0.25,"discount":0.5,"capabilities":["Reasoning","Tools","Files","Vision"],"description":"Fastest Claude model for high-volume, latency-sensitive work such as classification, extraction, and routing"},{"id":"claude-opus-5","family":"claude","family_label":"claude文本分组","endpoints":["anthropic","openai"],"context_length":1000000,"max_output_tokens":128000,"input_usd_per_million":2.5,"output_usd_per_million":12.5,"discount":0.5,"capabilities":["Reasoning","Tools","Files","Vision"],"description":"Strongest Claude Opus model for coding, agents, and professional work"},{"id":"claude-opus-5-5","family":"claude","family_label":"claude文本分组","endpoints":["anthropic","openai"],"context_length":1000000,"max_output_tokens":128000,"input_usd_per_million":2,"output_usd_per_million":10,"discount":0.5,"capabilities":["Reasoning","Tools","Files","Vision"],"description":"Claude Opus model for long-running agentic coding and knowledge work, priced below Opus 5"},{"id":"claude-sonnet-5","family":"claude","family_label":"claude文本分组","endpoints":["anthropic","openai"],"context_length":1000000,"max_output_tokens":128000,"input_usd_per_million":1,"output_usd_per_million":5,"discount":0.5,"capabilities":["Reasoning","Tools","Files","Vision"],"description":"Everyday Claude agent model for coding, planning, browsing, and general work"},{"id":"claude-sonnet-5-5","family":"claude","family_label":"claude文本分组","endpoints":["anthropic","openai"],"context_length":1000000,"max_output_tokens":128000,"input_usd_per_million":1,"output_usd_per_million":5,"discount":0.5,"capabilities":["Reasoning","Tools","Vision","Coding","编程","推理"],"description":"Claude Sonnet 5.5 for coding, reasoning, and professional work with a 1M-token context window."},{"id":"gpt-5.6-luna","family":"codex","family_label":"codex文本分组","endpoints":["openai"],"context_length":1050000,"max_output_tokens":128000,"input_usd_per_million":0.08000000000000002,"output_usd_per_million":0.48,"discount":0.4,"capabilities":["Reasoning","Tools","Files","Vision"],"description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads"},{"id":"gpt-5.6-sol","family":"codex","family_label":"codex文本分组","endpoints":["openai"],"context_length":1050000,"max_output_tokens":128000,"input_usd_per_million":2,"output_usd_per_million":12,"discount":0.4,"capabilities":["Reasoning","Tools","Files","Vision"],"description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows"},{"id":"gpt-5.6-terra","family":"codex","family_label":"codex文本分组","endpoints":["openai"],"context_length":1050000,"max_output_tokens":128000,"input_usd_per_million":1,"output_usd_per_million":6,"discount":0.4,"capabilities":["Reasoning","Tools","Files","Vision"],"description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work"},{"id":"gpt-6-astra","family":"codex","family_label":"codex文本分组","endpoints":["openai"],"context_length":1000000,"max_output_tokens":128000,"input_usd_per_million":4,"output_usd_per_million":20,"discount":0.4,"capabilities":["推理","工具","文件","多模态"],"description":"GPT-6 Astra 是 OpenAI 能力最强的模型，面向复杂推理、编程、计算机操作、研究与文档创作。"},{"id":"gpt-6-luna","family":"codex","family_label":"codex文本分组","endpoints":["openai"],"context_length":1050000,"max_output_tokens":128000,"input_usd_per_million":0.04000000000000001,"output_usd_per_million":0.2,"discount":0.4,"capabilities":["Reasoning","Tools","Files","Vision"],"description":"Cost-efficient GPT-6 model for focused, high-volume tasks"},{"id":"gpt-6-sol","family":"codex","family_label":"codex文本分组","endpoints":["openai"],"context_length":1050000,"max_output_tokens":128000,"input_usd_per_million":0.8,"output_usd_per_million":4,"discount":0.4,"capabilities":["Reasoning","Tools","Files","Vision"],"description":"Mid-tier GPT-6 model for everyday coding, research, and agentic work at a lower price"},{"id":"gpt-6.1-sol","family":"codex","family_label":"codex文本分组","endpoints":["openai","openai-response"],"context_length":1050000,"max_output_tokens":128000,"input_usd_per_million":0.8,"output_usd_per_million":4,"discount":0.4,"capabilities":["Reasoning","Tools","Vision","Coding","代码","推理"],"description":"GPT-6.1 Sol for coding, computer use, and professional work. Use Responses for tool calling."},{"id":"gpt-6.1-terra","family":"codex","family_label":"codex文本分组","endpoints":["openai","openai-response"],"context_length":1050000,"max_output_tokens":128000,"input_usd_per_million":0.4,"output_usd_per_million":2,"discount":0.4,"capabilities":["Reasoning","Tools","Vision","Coding","代码","推理"],"description":"GPT-6.1 Terra for visual reasoning, computer use, and writing."},{"id":"deepseek-v4-flash-0731","family":"deepseek","family_label":"deepseek文本分组","endpoints":["openai"],"context_length":1000000,"max_output_tokens":128000,"input_usd_per_million":0.11,"output_usd_per_million":0.33,"discount":0.5},{"id":"deepseek-v4-pro-0813","family":"deepseek","family_label":"deepseek文本分组","endpoints":["openai"],"context_length":1000000,"max_output_tokens":128000,"input_usd_per_million":0.33,"output_usd_per_million":0.99,"discount":0.5},{"id":"deepseek-v4.1-flash","family":"deepseek","family_label":"deepseek文本分组","endpoints":["openai"],"input_usd_per_million":0.075,"output_usd_per_million":0.3,"discount":0.5},{"id":"deepseek-v4-pro-0813-hc","family":"deepseek-hc","family_label":"deepseek高缓存文本分组","endpoints":["openai"],"context_length":1000000,"max_output_tokens":128000,"input_usd_per_million":0.46199999999999997,"output_usd_per_million":1.386,"time_rules":[{"multiplier":2,"when":"weekday(\"Asia/Shanghai\") \u003e= 1 \u0026\u0026 weekday(\"Asia/Shanghai\") \u003c= 5 \u0026\u0026 ((hour(\"Asia/Shanghai\") \u003e= 9 \u0026\u0026 hour(\"Asia/Shanghai\") \u003c 12) || (hour(\"Asia/Shanghai\") \u003e= 14 \u0026\u0026 hour(\"Asia/Shanghai\") \u003c 18))","active":false}],"discount":0.7},{"id":"deepseek-v4.1-flash-hc","family":"deepseek-hc","family_label":"deepseek高缓存文本分组","endpoints":["openai"],"input_usd_per_million":0.105,"output_usd_per_million":0.42,"time_rules":[{"multiplier":2,"when":"weekday(\"Asia/Shanghai\") \u003e= 1 \u0026\u0026 weekday(\"Asia/Shanghai\") \u003c= 5 \u0026\u0026 ((hour(\"Asia/Shanghai\") \u003e= 9 \u0026\u0026 hour(\"Asia/Shanghai\") \u003c 12) || (hour(\"Asia/Shanghai\") \u003e= 14 \u0026\u0026 hour(\"Asia/Shanghai\") \u003c 18))","active":false}],"discount":0.7},{"id":"deepseek-v4.1-flash-free","family":"free","family_label":"免费模型分组（随时可能下架）","endpoints":["openai"]},{"id":"glm-5.3-flash-free","family":"free","family_label":"免费模型分组（随时可能下架）","endpoints":["openai"]},{"id":"hy4-free","family":"free","family_label":"免费模型分组（随时可能下架）","endpoints":["openai"]},{"id":"gemini-3.1-pro-preview","family":"gemini","family_label":"gemini文本分组","endpoints":["openai"],"context_length":1048576,"max_output_tokens":65536,"input_usd_per_million":1,"output_usd_per_million":6,"discount":0.5,"capabilities":["Reasoning","Tools","Files","Vision","Audio"],"description":"Reasoning-first Gemini preview for agentic coding and complex problem solving"},{"id":"gemini-3.6-flash","family":"gemini","family_label":"gemini文本分组","endpoints":["openai"],"context_length":1048576,"max_output_tokens":128000,"input_usd_per_million":0.375,"output_usd_per_million":1.875,"discount":0.5,"capabilities":["Reasoning","Tools","Files","Vision","Audio"],"description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost"},{"id":"gemini-3.7-flash","family":"gemini","family_label":"gemini文本分组","endpoints":["openai"],"context_length":1048576,"max_output_tokens":128000,"input_usd_per_million":0.375,"output_usd_per_million":1.875,"discount":0.5,"capabilities":["Reasoning","Tools","Files","Vision","Audio"],"description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning"},{"id":"gemini-3.8-flash","family":"gemini","family_label":"gemini文本分组","endpoints":["openai"],"input_usd_per_million":0.375,"output_usd_per_million":1.875,"discount":0.5,"capabilities":["Reasoning","Tools","Files","Vision","Audio"],"description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows"},{"id":"gemini-4-argon","family":"gemini","family_label":"gemini文本分组","endpoints":["openai"],"max_output_tokens":1000000,"input_usd_per_million":1,"output_usd_per_million":5,"discount":0.5,"capabilities":["Reasoning","Tools","Vision"],"description":"Google's Gemini 4 flagship model, with an output limit of up to 1M tokens."},{"id":"gemini-4-carbon","family":"gemini","family_label":"gemini文本分组","endpoints":["openai"],"max_output_tokens":1000000,"input_usd_per_million":1,"output_usd_per_million":5,"discount":0.5,"capabilities":["Reasoning","Tools","Vision"],"description":"Google's Gemini 4 Carbon model, a newer Gemini 4 text model with reasoning, tools and vision input."},{"id":"glm-5.3","family":"glm","family_label":"glm文本分组","endpoints":["openai"],"context_length":1000000,"max_output_tokens":128000,"input_usd_per_million":0.5599999999999999,"output_usd_per_million":1.7600000000000002,"discount":0.4,"capabilities":["推理","工具","开源权重"],"description":"GLM 旗舰模型，面向长程编程、智能体与复杂项目交付"},{"id":"glm-5.3-flash","family":"glm","family_label":"glm文本分组","endpoints":["openai"],"context_length":1000000,"max_output_tokens":128000,"input_usd_per_million":0.06,"output_usd_per_million":0.2,"discount":0.4,"capabilities":["推理","工具","文件","开源权重","多模态"],"description":"原生多模态 GLM 模型，面向高效编程与长程智能体任务"},{"id":"glm-5.3-flashx","family":"glm","family_label":"glm文本分组","endpoints":["openai"],"context_length":1000000,"max_output_tokens":128000,"input_usd_per_million":0.148,"output_usd_per_million":0.5,"discount":0.4,"capabilities":["推理","工具","文件","开源权重","多模态"],"description":"GLM-5.3-Flash 的高速服务版本，面向编程与智能体工作流"},{"id":"grok-4.5","family":"grok","family_label":"grok文本分组","endpoints":["openai"],"context_length":500000,"max_output_tokens":128000,"input_usd_per_million":0.7,"output_usd_per_million":2.0999999999999996,"discount":0.35,"capabilities":["Reasoning","Tools","Files","Vision"],"description":"xAI's Grok model for chat, coding, agentic tools, and lower hallucination risk"},{"id":"grok-4.6","family":"grok","family_label":"grok文本分组","endpoints":["openai"],"context_length":500000,"max_output_tokens":128000,"input_usd_per_million":0.7,"output_usd_per_million":2.0999999999999996,"discount":0.35,"capabilities":["Reasoning","Tools","Files","Vision"],"description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects"},{"id":"grok-4.7","family":"grok","family_label":"grok文本分组","endpoints":["openai"],"context_length":500000,"input_usd_per_million":0.7,"output_usd_per_million":2.0999999999999996,"discount":0.35,"capabilities":["Reasoning","Tools","Files","Vision"],"description":"xAI's most capable Grok for coding, agents, and knowledge work; works longer on hard tasks and checks its own work"},{"id":"hy3","family":"hunyuan","family_label":"hunyuan文本分组","endpoints":["openai"],"context_length":256000,"max_output_tokens":64000,"input_usd_per_million":0.033,"output_usd_per_million":0.132,"time_rules":[{"multiplier":1.6,"when":"hour(\"Asia/Shanghai\") \u003e= 8","active":false}],"discount":0.4,"capabilities":["推理","工具","开源权重"],"description":"腾讯 Hy 推理模型，面向编程、指令遵循与智能体任务"},{"id":"hy4-preview","family":"hunyuan","family_label":"hunyuan文本分组","endpoints":["openai"],"input_usd_per_million":0.3336,"output_usd_per_million":1.0004,"discount":0.4,"capabilities":["Reasoning","Tools","Open Weights"],"description":"A next-generation productivity model with significantly enhanced Agent and complex task execution capabilities."},{"id":"kimi-k3","family":"kimi","family_label":"kimi文本分组","endpoints":["openai"],"context_length":1048576,"max_output_tokens":128000,"input_usd_per_million":1.2000000000000002,"output_usd_per_million":6,"discount":0.4,"capabilities":["推理","工具","文件","开源权重","多模态"],"description":"多模态 Kimi 模型，具备 100 万 token 上下文与可切换的最高强度思考模式，面向长程智能体任务"},{"id":"kimi-k3.1-pro","family":"kimi","family_label":"kimi文本分组","endpoints":["openai"],"context_length":1048576,"input_usd_per_million":0.8,"output_usd_per_million":4,"discount":0.4,"capabilities":["推理","工具"],"description":"Kimi K3.1 Pro，100 万 token 上下文的 Kimi 旗舰模型，面向长程智能体、推理与编程任务"},{"id":"mimo-v2.5","family":"mimo","family_label":"mimo文本分组","endpoints":["openai"],"context_length":1048576,"max_output_tokens":128000,"input_usd_per_million":0.05600000000000001,"output_usd_per_million":0.11200000000000002,"discount":0.4,"capabilities":["推理","工具","文件","开源权重","多模态","音频"],"description":"开源 MiMo 模型，面向多模态编程智能体与长上下文自动化"},{"id":"mimo-v2.6-flash","family":"mimo","family_label":"mimo文本分组","endpoints":["openai"],"context_length":1048576,"max_output_tokens":128000,"input_usd_per_million":0.05600000000000001,"output_usd_per_million":0.11200000000000002,"discount":0.4,"capabilities":["推理","工具","文件","开源权重","多模态","音频"],"description":"MiMo Flash model for multimodal coding agents and long-context automation"},{"id":"MiniMax-M3","family":"minimax","family_label":"minimax文本分组","endpoints":["openai"],"context_length":1048576,"max_output_tokens":128000,"input_usd_per_million":0.12,"output_usd_per_million":0.48,"discount":0.4,"capabilities":["推理","工具","文件","开源权重","多模态"],"description":"MiniMax 多模态模型，面向长上下文编程、感知与智能体规划"},{"id":"MiniMax-M3.1-Flash","family":"minimax","family_label":"minimax文本分组","endpoints":["openai"],"context_length":1000000,"input_usd_per_million":0.06,"output_usd_per_million":0.24,"discount":0.4,"capabilities":["推理","工具","多模态"],"description":"MiniMax M3.1 Flash，面向智能体推理、工具调用、编程与长上下文的高性价比模型"},{"id":"muse-wonder","family":"muse","family_label":"muse文本分组","endpoints":["openai"],"context_length":1048576,"max_output_tokens":131072,"input_usd_per_million":4,"output_usd_per_million":20,"capabilities":["Reasoning","Tools","Vision"],"description":"Meta Muse Wonder: a frontier reasoning model for complex personal and professional agentic work."},{"id":"qwen3.8-flash","family":"qwen","family_label":"qwen文本分组","endpoints":["openai"],"input_usd_per_million":0.06,"output_usd_per_million":0.188,"discount":0.4,"capabilities":["Reasoning","Tools","Files","Vision"],"description":"Qwen vision-language model for visual reasoning, documents, and agent tasks"},{"id":"qwen3.8-max","family":"qwen","family_label":"qwen文本分组","endpoints":["openai"],"context_length":1000000,"max_output_tokens":128000,"input_usd_per_million":0.8,"output_usd_per_million":2.4000000000000004,"discount":0.4,"capabilities":["推理","工具","文件","多模态"],"description":"2.4 万亿参数的 MoE 旗舰模型，面向编程、专业工作、多模态理解与长程智能体工作流"}]}}