{ "openapi": "3.1.0", "info": { "title": "MiniMax Text API Anthropic", "description": "MiniMax 文本生成 API，支持对话补全与流式输出。\n\n> ⚡ **`MiniMax-M3.1-Flash-Preview` 亮点** —— **1M 超长上下文**、**多模态**、**思考深度可调**。\n", "license": { "name": "MIT" }, "version": "1.0.0" }, "servers": [ { "url": "https://api.minimax.cn" } ], "security": [ { "bearerAuth": [] }, { "apiKeyAuth": [] } ], "paths": { "/anthropic/v1/messages": { "post": { "summary": "Text Generation Anthropic", "operationId": "chatCompletionAnthropic", "tags": [ "Text Generation" ], "parameters": [ { "name": "Content-Type", "in": "header", "required": true, "description": "请求体的媒介类型，请设置为 `application/json`，确保请求数据的格式为 JSON", "schema": { "type": "string", "enum": [ "application/json" ], "default": "application/json" } } ], "requestBody": { "content": { "application/json": { "schema": { "$ref": "#/components/schemas/CreateMessageReq" }, "examples": { "深度思考": { "value": { "model": "MiniMax-M3.1-Flash-Preview", "thinking": { "type": "adaptive" }, "messages": [ { "role": "user", "content": "9.11 和 9.9 哪个更大？" } ], "max_tokens": 4096, "output_config": { "effort": "max" } } }, "图片理解": { "value": { "model": "MiniMax-M3.1-Flash-Preview", "messages": [ { "role": "user", "content": [ { "type": "text", "text": "这张图片的内容是什么？" }, { "type": "image", "source": { "type": "url", "url": "https://filecdn.minimax.chat/public/fe9d04da-f60e-444d-a2e0-18ae743add33.jpeg" } } ] } ], "max_tokens": 4096, "thinking": { "type": "adaptive" }, "output_config": { "effort": "max" } } }, "视频理解": { "value": { "model": "MiniMax-M3.1-Flash-Preview", "thinking": { "type": "adaptive" }, "messages": [ { "role": "user", "content": [ { "type": "text", "text": "这个视频里发生了什么？" }, { "type": "video", "source": { "type": "url", "url": "https://filecdn.minimax.chat/public/ee8c1648-21f1-41b7-8397-65022d22ffe5.mp4" } } ] } ], "max_tokens": 4096, "output_config": { "effort": "max" } } }, "流式": { "value": { "model": "MiniMax-M3.1-Flash-Preview", "messages": [ { "role": "user", "content": [ { "type": "text", "text": "这张图片的内容是什么？" }, { "type": "image", "source": { "type": "url", "url": "https://filecdn.minimax.chat/public/fe9d04da-f60e-444d-a2e0-18ae743add33.jpeg" } } ] } ], "stream": true, "max_tokens": 4096, "thinking": { "type": "adaptive" }, "output_config": { "effort": "max" } } }, "工具调用": { "value": { "model": "MiniMax-M3.1-Flash-Preview", "messages": [ { "role": "user", "content": "旧金山现在天气怎么样？" } ], "max_tokens": 4096, "tools": [ { "name": "get_weather", "description": "Get the current weather for a given location.", "input_schema": { "type": "object", "properties": { "location": { "type": "string", "description": "The city and state/country, e.g. San Francisco, US" } }, "required": [ "location" ] } } ], "tool_choice": { "type": "auto" }, "output_config": { "effort": "max" } } } } } }, "required": true }, "responses": { "200": { "description": "", "content": { "application/json": { "schema": { "$ref": "#/components/schemas/CreateMessageResp" }, "examples": { "深度思考": { "value": { "id": "066b367547c2650d17dc215f503da551", "type": "message", "role": "assistant", "model": "MiniMax-M3.1-Flash-Preview", "content": [ { "thinking": "The user is asking which is larger: 9.11 or 9.9.\n\nComparing 9.11 and 9.9:\n9.9 = 9.90\n9.11 = 9.11\n\n9.90 > 9.11, so 9.9 is larger.", "signature": "7564f4e0e54b5c08b380d0b800aeb5463ea050e546cec7722e08ef6e912f6c67", "type": "thinking" }, { "text": "**9.9 更大。**\n\n比较方法很简单：将两个数的小数位数对齐后比较：\n\n- 9.11 = 9.**11**\n- 9.9 = 9.**90**\n\n因为 90 > 11，所以 **9.9 > 9.11**。\n\n这是一个常见的思维陷阱——虽然 11 看起来比 9 大，但在比较小数时，应该先比较整数部分（都是 9），然后再比较小数部分，而小数部分需要**补齐位数**后再比较。", "type": "text" } ], "usage": { "input_tokens": 13, "output_tokens": 172, "cache_creation_input_tokens": 0, "cache_read_input_tokens": 157 }, "stop_reason": "end_turn" } }, "图片理解": { "value": { "id": "066a381bdc3c0ded310e27c9a46d16e7", "type": "message", "role": "assistant", "model": "MiniMax-M3.1-Flash-Preview", "content": [ { "thinking": "The user is asking in Chinese what the content of the image is. I should describe the image in Chinese.", "signature": "6b19dfedcc8d3e065eb0ebe4262efd4a69e3e73ad69eeb7e5e629e552c40f969", "type": "thinking" }, { "text": "这张图片是一张温馨的儿童肖像照片，内容如下：\n\n**主体**：一个大约4-6岁的小女孩，有着棕色的卷曲长发和齐刘海，扎着蓬松的发型。她有一双大大的浅棕色（琥珀色）眼睛，长长的睫毛，粉嫩的嘴唇微微上扬，露出甜美自然的微笑。\n\n**服饰**：她穿着一件米白色或浅奶油色的连衣裙，领口和肩部有精致的蕾丝花边装饰，袖子是蓬松的荷叶边设计，整体风格显得优雅可爱。\n\n**光影与背景**：照片采用了柔和的暖色调光线，背景是朦胧的米灰色/暖灰色，营造出温馨柔和的氛围。光线从侧面打来，突出了她脸部的轮廓和头发的质感。\n\n**整体风格**：这是一张高质量的人像摄影作品，采用了浅景深的虚化效果（背景模糊），突出了小女孩纯真可爱的形象，画面充满童真与温暖感。", "type": "text" } ], "usage": { "input_tokens": 1209, "output_tokens": 211, "cache_creation_input_tokens": 0, "cache_read_input_tokens": 156 }, "stop_reason": "end_turn" } }, "视频理解": { "value": { "id": "066b4573467fa2bb299cee687b68873d", "type": "message", "role": "assistant", "model": "MiniMax-M3.1-Flash-Preview", "content": [ { "thinking": "The user is asking in Chinese: \"What happened in this video?\"\n\nI need to describe the events in the video clip provided.\n\n1. **Analyze the initial state (0.0s - 0.2s):** A spotted cat, resembling a Bengal cat, is walking confidently on a springboard, which looks like a diving board. The background shows an Olympic setting (\"Lausanne 2022\" and the Olympic rings are visible). There's a crowd watching in the background. The cat is moving towards the end of the board.\n2. **Observe the action (0.2s - 0.4s):** The cat reaches the end of the springboard and leaps off.\n3. **Follow the trajectory (0.4s - 0.5s):** The cat is in mid-air, flying over a swimming pool. It looks like it's doing a dive.\n4. **Analyze the final state (0.5s - 0.6s):** The cat splashes into the blue water of the swimming pool, creating a white splash. A person standing by the side of the pool appears to be reacting or clapping.\n\n**Synthesize the description:**\n- Start with the subject: A cat.\n- Describe its initial action: Walking on a diving board (springboard).\n- Describe the setting: It looks like an Olympic swimming event (Lausanne 2022).\n- Describe the main event: The cat jumps off the diving board into the air.\n- Describe the conclusion: The cat lands in the swimming pool, making a splash.\n\n**Translate to Chinese (as requested implicitly by the prompt language, but I should respond in the language the user used. The user asked in Chinese, so I will respond in Chinese. Wait, let me double-check standard practices. Usually, I should answer in the same language as the prompt. Prompt is Chinese. So, answer in Chinese.)\n\n**Drafting the Chinese response:**\n- 视频里，一只猫正在走路。\n- 它走到跳板的尽头。\n- 然后它从跳板上跳了下去。\n- 它在空中飞了一会儿。\n- 最后，它一头扎进了游泳池里，溅起了一片水花。\n- 背景看起来像是在洛桑2022年奥运会。\n\nLet's refine it into a concise and accurate description.\n\"视频显示，一只猫在跳水板上行走，然后从跳水板上跳下，扎进游泳池里。\" (The video shows a cat walking on a diving board, then jumping off it and plunging into a swimming pool.)\n\nLet's add a bit more detail based on the visual evidence.\n\"在洛桑2022年奥运会的背景下，一只猫在跳板上走到尽头，然后纵身一跃，跳入泳池中。\" (Against the backdrop of the Lausanne 2022 Olympics, a cat walks to the end of a diving board, then leaps and jumps into a swimming pool.)\n\nLet's stick to a simple, direct description of the main action.\n\"一只猫在跳板上行走，随后从跳板上跳下，扎进了身后的游泳池中。\" (A cat walks on a diving board, then jumps off the board and plunges into the swimming pool behind it.)\n\nThis is a good, concise summary of the key events. It covers the initial state, the action, and the final state.", "signature": "3ebe06a965f7edbb379552ee8f6485100f473eb42f12fb3541d952d5f5908d46", "type": "thinking" }, { "text": "视频显示，一只猫在跳板上行走，随后从跳板上跳下，扎进了身后的游泳池中。", "type": "text" } ], "usage": { "input_tokens": 349, "output_tokens": 686, "cache_creation_input_tokens": 0, "cache_read_input_tokens": 1894 }, "stop_reason": "end_turn" } }, "流式": { "value": [ { "type": "message_start", "message": { "id": "066a38a98a4ffbf5554a468fe0cbedf6", "type": "message", "role": "assistant", "content": [], "model": "MiniMax-M3.1-Flash-Preview", "stop_reason": null, "stop_sequence": null, "usage": { "input_tokens": 0, "output_tokens": 0, "cache_creation_input_tokens": 0, "cache_read_input_tokens": 1365 }, "service_tier": "standard" } }, { "type": "ping" }, { "type": "content_block_start", "index": 0, "content_block": { "type": "thinking", "thinking": "" } }, { "type": "content_block_delta", "index": 0, "delta": { "type": "thinking_delta", "thinking": "The user" } }, { "type": "content_block_delta", "index": 0, "delta": { "type": "thinking_delta", "thinking": " is asking in Chinese what the content of this image is. Let me describe it in Chinese.\n\nThe" } }, { "type": "content_block_delta", "index": 0, "delta": { "type": "thinking_delta", "thinking": " image shows a close-up portrait of a young girl with curly/wavy brown hair and bangs." } }, { "type": "content_block_delta", "index": 0, "delta": { "type": "thinking_delta", "thinking": " She has brown eyes and is smiling slightly at the camera. She's wearing a cream/iv" } }, { "type": "content_block_delta", "index": 0, "delta": { "type": "thinking_delta", "thinking": "ory colored dress with lace details and ruffles. The lighting is soft and warm, creating a classic" } }, { "type": "content_block_delta", "index": 0, "delta": { "type": "thinking_delta", "thinking": " portrait photography feel. The background is a neutral olive/grayish color, blurred to keep focus" } }, { "type": "content_block_delta", "index": 0, "delta": { "type": "thinking_delta", "thinking": " on the subject." } }, { "type": "content_block_delta", "index": 0, "delta": { "type": "signature_delta", "signature": "6deea14a039646c64cff2cdc13db8e1aa093c3f7334bfd4941ee1eee8663b214" } }, { "type": "content_block_stop", "index": 0 }, { "type": "content_block_start", "index": 1, "content_block": { "type": "text", "text": "" } }, { "type": "content_block_delta", "index": 1, "delta": { "type": "text_delta", "text": "这张图片是一张温馨的儿童人像摄影作品，主要内容如下" } }, { "type": "content_block_delta", "index": 1, "delta": { "type": "text_delta", "text": "：\n\n**人物：**\n- 一个大约5-7岁的小女孩\n- 有着卷曲蓬" } }, { "type": "content_block_delta", "index": 1, "delta": { "type": "text_delta", "text": "松的棕色头发，留着刘海\n- 有着明亮的棕色大眼睛\n- 微微" } }, { "type": "content_block_delta", "index": 1, "delta": { "type": "text_delta", "text": "带着甜美的微笑看向镜头\n\n**服装：**\n- 身穿一件米白色/象牙色的" } }, { "type": "content_block_delta", "index": 1, "delta": { "type": "text_delta", "text": "连衣裙\n- 衣领和肩部有精致的蕾丝花边和褶" } }, { "type": "content_block_delta", "index": 1, "delta": { "type": "text_delta", "text": "皱装饰\n\n**拍摄风格：**\n- 经典的近距离肖像构图\n- 柔和的暖色调" } }, { "type": "content_block_delta", "index": 1, "delta": { "type": "text_delta", "text": "光线，营造出温馨的氛围\n- 背景是模糊的橄榄灰色调，使人物更加突出" } }, { "type": "content_block_delta", "index": 1, "delta": { "type": "text_delta", "text": "\n- 整体呈现出专业的人像摄影质感，光影处理细腻\n\n这是一张" } }, { "type": "content_block_delta", "index": 1, "delta": { "type": "text_delta", "text": "很有艺术感的儿童写真照，捕捉到了小女孩天真可爱的瞬间。" } }, { "type": "content_block_stop", "index": 1 }, { "type": "message_delta", "delta": { "stop_reason": "end_turn" }, "usage": { "input_tokens": 1251, "output_tokens": 264, "cache_creation_input_tokens": 0, "cache_read_input_tokens": 114 } }, { "type": "message_stop" } ] }, "工具调用": { "value": { "id": "066b13de1224099141792cd806ed4f56", "type": "message", "role": "assistant", "model": "MiniMax-M3.1-Flash-Preview", "content": [ { "thinking": "The user is asking about the current weather in San Francisco. I should use the get_weather tool to fetch this information. The location parameter requires city and state/country format, so \"San Francisco, US\" would be appropriate.", "signature": "ca8010ef1c5606561a17158c090bd20f33a8a8ac5f3b6cfb4326408395cbff39", "type": "thinking" }, { "type": "tool_use", "id": "call_function_debp7sfnk20c_1", "name": "get_weather", "input": { "location": "San Francisco, US" } } ], "usage": { "input_tokens": 14, "output_tokens": 79, "cache_creation_input_tokens": 0, "cache_read_input_tokens": 397 }, "stop_reason": "tool_use" } } } }, "text/event-stream": { "schema": { "$ref": "#/components/schemas/StreamEvent" }, "examples": { "流式": { "value": [ { "type": "message_start", "message": { "id": "066a38a98a4ffbf5554a468fe0cbedf6", "type": "message", "role": "assistant", "content": [], "model": "MiniMax-M3.1-Flash-Preview", "stop_reason": null, "stop_sequence": null, "usage": { "input_tokens": 0, "output_tokens": 0, "cache_creation_input_tokens": 0, "cache_read_input_tokens": 1365 }, "service_tier": "standard" } }, { "type": "ping" }, { "type": "content_block_start", "index": 0, "content_block": { "type": "thinking", "thinking": "" } }, { "type": "content_block_delta", "index": 0, "delta": { "type": "thinking_delta", "thinking": "The user" } }, { "type": "content_block_delta", "index": 0, "delta": { "type": "thinking_delta", "thinking": " is asking in Chinese what the content of this image is. Let me describe it in Chinese.\n\nThe" } }, { "type": "content_block_delta", "index": 0, "delta": { "type": "thinking_delta", "thinking": " image shows a close-up portrait of a young girl with curly/wavy brown hair and bangs." } }, { "type": "content_block_delta", "index": 0, "delta": { "type": "thinking_delta", "thinking": " She has brown eyes and is smiling slightly at the camera. She's wearing a cream/iv" } }, { "type": "content_block_delta", "index": 0, "delta": { "type": "thinking_delta", "thinking": "ory colored dress with lace details and ruffles. The lighting is soft and warm, creating a classic" } }, { "type": "content_block_delta", "index": 0, "delta": { "type": "thinking_delta", "thinking": " portrait photography feel. The background is a neutral olive/grayish color, blurred to keep focus" } }, { "type": "content_block_delta", "index": 0, "delta": { "type": "thinking_delta", "thinking": " on the subject." } }, { "type": "content_block_delta", "index": 0, "delta": { "type": "signature_delta", "signature": "6deea14a039646c64cff2cdc13db8e1aa093c3f7334bfd4941ee1eee8663b214" } }, { "type": "content_block_stop", "index": 0 }, { "type": "content_block_start", "index": 1, "content_block": { "type": "text", "text": "" } }, { "type": "content_block_delta", "index": 1, "delta": { "type": "text_delta", "text": "这张图片是一张温馨的儿童人像摄影作品，主要内容如下" } }, { "type": "content_block_delta", "index": 1, "delta": { "type": "text_delta", "text": "：\n\n**人物：**\n- 一个大约5-7岁的小女孩\n- 有着卷曲蓬" } }, { "type": "content_block_delta", "index": 1, "delta": { "type": "text_delta", "text": "松的棕色头发，留着刘海\n- 有着明亮的棕色大眼睛\n- 微微" } }, { "type": "content_block_delta", "index": 1, "delta": { "type": "text_delta", "text": "带着甜美的微笑看向镜头\n\n**服装：**\n- 身穿一件米白色/象牙色的" } }, { "type": "content_block_delta", "index": 1, "delta": { "type": "text_delta", "text": "连衣裙\n- 衣领和肩部有精致的蕾丝花边和褶" } }, { "type": "content_block_delta", "index": 1, "delta": { "type": "text_delta", "text": "皱装饰\n\n**拍摄风格：**\n- 经典的近距离肖像构图\n- 柔和的暖色调" } }, { "type": "content_block_delta", "index": 1, "delta": { "type": "text_delta", "text": "光线，营造出温馨的氛围\n- 背景是模糊的橄榄灰色调，使人物更加突出" } }, { "type": "content_block_delta", "index": 1, "delta": { "type": "text_delta", "text": "\n- 整体呈现出专业的人像摄影质感，光影处理细腻\n\n这是一张" } }, { "type": "content_block_delta", "index": 1, "delta": { "type": "text_delta", "text": "很有艺术感的儿童写真照，捕捉到了小女孩天真可爱的瞬间。" } }, { "type": "content_block_stop", "index": 1 }, { "type": "message_delta", "delta": { "stop_reason": "end_turn" }, "usage": { "input_tokens": 1251, "output_tokens": 264, "cache_creation_input_tokens": 0, "cache_read_input_tokens": 114 } }, { "type": "message_stop" } ] } } } } }, "400": { "description": "请求参数非法（必填缺失、type 不在白名单、tool_use.input 非 JSON 对象等）", "content": { "application/json": { "schema": { "$ref": "#/components/schemas/ErrorResponse" }, "example": { "type": "error", "request_id": "req_xxxxxxxx", "error": { "type": "invalid_request_error", "message": "messages.0.content.1: unsupported content type 'foo'" } } } } }, "401": { "description": "API Key 缺失/无效", "content": { "application/json": { "schema": { "$ref": "#/components/schemas/ErrorResponse" }, "example": { "type": "error", "request_id": "req_xxxxxxxx", "error": { "type": "authentication_error", "message": "API Key 缺失/无效" } } } } }, "403": { "description": "无权访问该模型或该路径", "content": { "application/json": { "schema": { "$ref": "#/components/schemas/ErrorResponse" }, "example": { "type": "error", "request_id": "req_xxxxxxxx", "error": { "type": "permission_error", "message": "无权访问该模型或该路径" } } } } }, "404": { "description": "模型不存在", "content": { "application/json": { "schema": { "$ref": "#/components/schemas/ErrorResponse" }, "example": { "type": "error", "request_id": "req_xxxxxxxx", "error": { "type": "not_found_error", "message": "模型不存在" } } } } }, "413": { "description": "请求体超过 64MB，或多模态文件超出大小限制", "content": { "application/json": { "schema": { "$ref": "#/components/schemas/ErrorResponse" }, "example": { "type": "error", "request_id": "req_xxxxxxxx", "error": { "type": "request_too_large", "message": "请求体超过 64MB，或多模态文件超出大小限制" } } } } }, "429": { "description": "触发 RPM/TPM/连接数 等限流", "content": { "application/json": { "schema": { "$ref": "#/components/schemas/ErrorResponse" }, "example": { "type": "error", "request_id": "req_xxxxxxxx", "error": { "type": "rate_limit_error", "message": "触发 RPM/TPM/连接数 等限流" } } } } }, "500": { "description": "服务端内部错误", "content": { "application/json": { "schema": { "$ref": "#/components/schemas/ErrorResponse" }, "example": { "type": "error", "request_id": "req_xxxxxxxx", "error": { "type": "api_error", "message": "服务端内部错误" } } } } }, "529": { "description": "上游模型过载，可重试", "content": { "application/json": { "schema": { "$ref": "#/components/schemas/ErrorResponse" }, "example": { "type": "error", "request_id": "req_xxxxxxxx", "error": { "type": "overloaded_error", "message": "上游模型过载，可重试" } } } } } } } }, "/anthropic/v1/messages/count_tokens": { "post": { "summary": "Anthropic Token 计数", "operationId": "countTokensAnthropic", "tags": [ "Text Generation" ], "parameters": [ { "name": "Content-Type", "in": "header", "required": true, "description": "请求体的媒介类型，请设置为 `application/json`，确保请求数据的格式为 JSON", "schema": { "type": "string", "enum": [ "application/json" ], "default": "application/json" } } ], "requestBody": { "content": { "application/json": { "schema": { "$ref": "#/components/schemas/CreateMessageReq" }, "examples": { "标准": { "value": { "model": "MiniMax-M3.1-Flash-Preview", "messages": [ { "role": "user", "content": "Hello" } ] } } } } }, "required": true }, "responses": { "200": { "description": "返回本次请求的输入 token 用量，不生成模型输出。", "content": { "application/json": { "schema": { "$ref": "#/components/schemas/CountTokensResp" }, "examples": { "标准": { "value": { "usage": { "input_tokens": 3, "cache_creation_input_tokens": 0, "cache_read_input_tokens": 0 } } } } } } } } } } }, "components": { "schemas": { "CreateMessageReq": { "type": "object", "required": [ "model", "messages" ], "properties": { "model": { "type": "string", "description": "模型 ID。`MiniMax-M3.1-Flash-Preview` 和 `MiniMax-M3` 是多模态模型，原生支持文本、图片和视频输入，并兼容工具调用与 thinking 内容块；`MiniMax-M3.1-Flash-Preview` 还会强制开启 thinking，并支持通过 `output_config.effort` 调节思考深度。M2.7、M2.5、M2.1 和 M2 系列仅支持文本与工具调用，不支持图片和视频输入。", "enum": [ "MiniMax-M3.1-Flash-Preview", "MiniMax-M3", "MiniMax-M2.7", "MiniMax-M2.7-highspeed", "MiniMax-M2.5", "MiniMax-M2.5-highspeed", "MiniMax-M2.1", "MiniMax-M2.1-highspeed", "MiniMax-M2" ] }, "service_tier": { "type": "string", "description": "请求准入服务层级。支持的取值为 `standard` 和 `priority`。省略时默认使用 `standard`。`priority` 的[价格](https://platform.minimax.cn/docs/pricing/overview#models/text)为 `standard` 的 1.5 倍，并会确保请求获得优先准入，使其排在其他请求之前处理，从而带来更快响应并减少失败。", "enum": [ "standard", "priority" ], "default": "standard" }, "system": { "description": "设置模型角色与行为。", "oneOf": [ { "type": "string", "description": "纯文本系统提示词" }, { "type": "array", "description": "内容块数组格式的系统提示词。text 块可携带 cache_control。", "items": { "type": "object", "properties": { "type": { "type": "string", "enum": [ "text" ], "description": "内容块类型" }, "text": { "type": "string", "description": "文本内容" }, "cache_control": { "$ref": "#/components/schemas/CacheControl" } }, "required": [ "type", "text" ] } } ] }, "messages": { "type": "array", "description": "对话历史。`MiniMax-M3.1-Flash-Preview` 和 `MiniMax-M3` 支持文本、图片、视频、工具调用、工具结果和 thinking 内容块。M2.7、M2.5、M2.1 和 M2 系列仅支持文本与工具调用相关内容块，不支持图片和视频输入。", "items": { "$ref": "#/components/schemas/Message" } }, "stream": { "type": "boolean", "description": "是否使用流式传输，默认为 `false`。设置为 `true` 后，响应将分批返回", "default": false }, "max_tokens": { "type": "integer", "format": "int64", "description": "指定生成内容长度的上限（Token 数）。MiniMax-M3.1-Flash-Preview 和 MiniMax-M3 推荐值为 131072（128K），上限为 524288（512K）；其他模型推荐值为 65536（64K），上限为 204800（200K）。超过上限的内容会被截断。如果生成因 `length` 原因中断，请尝试调高此值", "minimum": 1 }, "temperature": { "type": "number", "format": "double", "description": "温度系数，影响输出随机性，取值范围 [0, 2]，默认值为 1。值越高，输出越随机；值越低，输出越确定。", "minimum": 0, "maximum": 2, "default": 1 }, "top_p": { "type": "number", "format": "double", "description": "核采样参数，取值范围 [0, 1]。MiniMax-M3.1-Flash-Preview 和 MiniMax-M3 默认值为 0.95，M2.x 系列模型默认值为 0.9。", "minimum": 0, "maximum": 1, "default": 0.95 }, "tools": { "type": "array", "description": "Anthropic 兼容工具调用的工具定义。", "items": { "$ref": "#/components/schemas/Tool" } }, "tool_choice": { "$ref": "#/components/schemas/ToolChoice" }, "thinking": { "type": "object", "description": "控制 thinking 行为。默认值随模型不同，因此未在 schema 层声明统一默认值。\n- `MiniMax-M3.1-Flash-Preview`：强制开启 thinking。若传入 `type` 只能为 `adaptive`；传入 `disabled` 会返回 HTTP 400。如需调节思考深度，请使用 `output_config.effort`。\n- `MiniMax-M3`：省略 `thinking` 时默认关闭；传入 `adaptive` 可开启并返回 thinking 块。\n- M2.x 模型：thinking 无法关闭，传入 `disabled` 会被接收但不生效。", "properties": { "type": { "type": "string", "enum": [ "disabled", "adaptive" ], "description": "thinking 控制类型。\n- `disabled`：关闭 `MiniMax-M3` 的 thinking 输出。`MiniMax-M3.1-Flash-Preview` 传入该值会返回 HTTP 400；M2.x 模型会接收但不生效。\n- `adaptive`：开启 thinking 输出并返回 thinking 块。这是 `MiniMax-M3.1-Flash-Preview` 唯一接受的取值。" } }, "required": [] }, "output_config": { "type": "object", "description": "输出配置。通过 `effort` 调节 `MiniMax-M3.1-Flash-Preview` 的思考深度。", "properties": { "effort": { "type": "string", "description": "`MiniMax-M3.1-Flash-Preview` 的思考深度档位，可取 `low`、`medium`、`high`、`xhigh`、`max`；省略时默认为 `max`。其他模型会忽略该字段。`MiniMax-M3.1-Flash-Preview` 不支持 `none`，传入会返回 HTTP 400。", "enum": [ "low", "medium", "high", "xhigh", "max" ] } }, "required": [] }, "metadata": { "type": "object", "description": "请求元信息。建议对 to-C 业务传入 user_id，便于按终端用户聚合限流和计费分析。", "properties": { "user_id": { "type": "string", "description": "终端用户 ID。" } } } } }, "CreateMessageResp": { "type": "object", "properties": { "id": { "type": "string", "description": "本次响应的唯一 ID" }, "type": { "type": "string", "description": "对象类型，固定为 `message`", "enum": [ "message" ] }, "role": { "type": "string", "description": "角色，固定为 `assistant`", "enum": [ "assistant" ] }, "model": { "type": "string", "description": "本次请求使用的模型 ID" }, "content": { "type": "array", "description": "响应内容块列表", "items": { "$ref": "#/components/schemas/ResponseContentBlock" } }, "stop_reason": { "type": "string", "description": "模型停止生成的原因：\n- end_turn：模型自然结束\n- max_tokens：达到 max_tokens 限制\n- tool_use：模型请求工具调用", "enum": [ "end_turn", "max_tokens", "tool_use" ] }, "usage": { "$ref": "#/components/schemas/Usage" } } }, "Message": { "type": "object", "required": [ "role", "content" ], "properties": { "role": { "type": "string", "enum": [ "user", "assistant", "user_system", "group", "sample_message_user", "sample_message_ai" ], "description": "消息发送方角色。`MiniMax-M3.1-Flash-Preview` 和 `MiniMax-M3` 使用 user / assistant 交替消息。" }, "content": { "description": "消息内容。`MiniMax-M3.1-Flash-Preview` 和 `MiniMax-M3` 支持文本、图片、视频、工具调用、工具结果和 thinking 内容块。", "oneOf": [ { "type": "string", "description": "纯文本消息" }, { "type": "array", "description": "消息内容。`MiniMax-M3.1-Flash-Preview` 和 `MiniMax-M3` 支持文本、图片、视频、工具调用、工具结果和 thinking 内容块。", "items": { "$ref": "#/components/schemas/RequestContentBlock" } } ] } } }, "RequestContentBlock": { "type": "object", "description": "请求消息中可用的内容块。text 与工具相关块所有模型均支持；image 与 video 块仅 `MiniMax-M3.1-Flash-Preview` 和 `MiniMax-M3` 支持。thinking 仅在多轮对话中将上一轮 assistant 输出原样回带时使用。", "required": [ "type" ], "properties": { "type": { "type": "string", "description": "内容块类型：\n- text：文本内容\n- image：图片输入，仅 `MiniMax-M3.1-Flash-Preview` 和 `MiniMax-M3` 支持\n- video：视频输入，仅 `MiniMax-M3.1-Flash-Preview` 和 `MiniMax-M3` 支持\n- tool_use：回带上一轮的 assistant 工具调用\n- tool_result：工具执行结果\n- thinking：回带上一轮的 assistant 思考内容\n- mid_conv_system：对话中途插入的系统指令", "enum": [ "text", "image", "video", "tool_use", "tool_result", "thinking", "mid_conv_system" ] }, "text": { "type": "string", "description": "文本内容，type=text 时使用。" }, "source": { "$ref": "#/components/schemas/MediaSource", "description": "图片或视频来源，type=image 或 type=video 时使用，仅 `MiniMax-M3.1-Flash-Preview` 和 `MiniMax-M3` 支持。" }, "id": { "type": "string", "description": "回带 type=tool_use 时使用的工具调用 ID。" }, "name": { "type": "string", "description": "回带 type=tool_use 时使用的工具名称。" }, "input": { "type": "object", "description": "回带 type=tool_use 时使用的工具输入。" }, "tool_use_id": { "type": "string", "description": "对应 tool_use 块的 ID，type=tool_result 时使用。" }, "content": { "oneOf": [ { "type": "string" }, { "type": "array", "items": { "$ref": "#/components/schemas/RequestContentBlock" } } ], "description": "工具执行结果，可为字符串或 text/image 内容块数组。" }, "thinking": { "type": "string", "description": "回带的 assistant 思考过程内容，type=thinking 时使用。" }, "signature": { "type": "string", "description": "回带的 thinking 内容签名，多轮续写时需要原样回带。" }, "cache_control": { "$ref": "#/components/schemas/CacheControl" } } }, "ResponseContentBlock": { "type": "object", "description": "响应消息中的内容块。模型输出 text、tool_use（调用工具时），以及 thinking（开启深度思考时）。响应中不会出现 image、video、tool_result 或 mid_conv_system 块。", "required": [ "type" ], "properties": { "type": { "type": "string", "description": "内容块类型：\n- text：文本内容\n- tool_use：模型工具调用\n- thinking：模型思考过程", "enum": [ "text", "tool_use", "thinking" ] }, "text": { "type": "string", "description": "文本内容，type=text 时使用。" }, "id": { "type": "string", "description": "工具调用 ID，type=tool_use 时返回。" }, "name": { "type": "string", "description": "工具名称，type=tool_use 时返回。" }, "input": { "type": "object", "description": "工具输入参数，type=tool_use 时返回。" }, "thinking": { "type": "string", "description": "模型思考过程内容，type=thinking 时使用。" }, "signature": { "type": "string", "description": "thinking 内容签名，多轮续写时需要原样回带。" } } }, "Usage": { "type": "object", "description": "本次请求的 token 用量，包含适用时的 prompt cache 用量。", "properties": { "input_tokens": { "type": "integer", "description": "输入消耗的 Token 数" }, "output_tokens": { "type": "integer", "description": "输出消耗的 Token 数" }, "cache_creation_input_tokens": { "type": "integer", "description": "创建 prompt cache 的输入 token 数。" }, "cache_read_input_tokens": { "type": "integer", "description": "命中 prompt cache 的输入 token 数。" } } }, "StreamEvent": { "type": "object", "description": "", "required": [ "type" ], "properties": { "type": { "type": "string", "description": "事件类型：\n- `message_start`: 消息开始，包含完整的消息元数据\n- `ping`: 心跳事件\n- `content_block_start`: 内容块开始\n- `content_block_delta`: 内容块增量更新\n- `content_block_stop`: 内容块结束\n- `message_delta`: 消息级别的增量更新（如 stop_reason）\n- `message_stop`: 消息结束", "enum": [ "message_start", "ping", "content_block_start", "content_block_delta", "content_block_stop", "message_delta", "message_stop" ] }, "message": { "type": "object", "description": "消息对象（`type` 为 `message_start` 时返回）", "properties": { "id": { "type": "string", "description": "消息的唯一 ID" }, "type": { "type": "string", "enum": [ "message" ] }, "role": { "type": "string", "enum": [ "assistant" ] }, "content": { "type": "array", "description": "内容块列表，初始为空数组", "items": { "$ref": "#/components/schemas/ResponseContentBlock" } }, "model": { "type": "string", "description": "模型 ID" }, "stop_reason": { "type": "string", "nullable": true, "description": "停止原因，流式开始时为 null" }, "stop_sequence": { "type": "string", "nullable": true, "description": "停止序列，流式开始时为 null" }, "usage": { "$ref": "#/components/schemas/Usage" } } }, "index": { "type": "integer", "description": "内容块的索引（`content_block_start`、`content_block_delta`、`content_block_stop` 时返回）" }, "content_block": { "$ref": "#/components/schemas/ResponseContentBlock", "description": "内容块对象（`type` 为 `content_block_start` 时返回）" }, "delta": { "type": "object", "description": "增量更新内容（`content_block_delta` 或 `message_delta` 时返回）", "properties": { "type": { "type": "string", "description": "增量类型，例如 text_delta、thinking_delta 或 signature_delta。", "enum": [ "text_delta", "thinking_delta", "signature_delta" ] }, "text": { "type": "string", "description": "增量文本内容" }, "stop_reason": { "type": "string", "description": "模型停止生成的原因。", "enum": [ "end_turn", "max_tokens", "tool_use" ] } } }, "usage": { "$ref": "#/components/schemas/Usage", "description": "Token 使用情况（`message_delta` 时返回）" } } }, "CacheControl": { "type": "object", "description": "提示词缓存标记。", "properties": { "type": { "type": "string", "enum": [ "ephemeral" ], "description": "提示词缓存标记。" } }, "required": [ "type" ] }, "MediaSource": { "type": "object", "description": "图片或视频内容块的媒体来源。\n\n**支持的图片格式**\n\n| 格式 | 常见拓展名 | MIME Type |\n| :-- | :-- | :-- |\n| JPEG | .jpg, .jpeg | image/jpeg |\n| PNG | .png | image/png |\n| GIF | .gif | image/gif |\n| WEBP | .webp | image/webp |\n\n**支持的视频格式**\n\n| 格式 | 常见拓展名 | MIME Type |\n| :-- | :-- | :-- |\n| MP4 | .mp4 | video/mp4 |\n| AVI | .avi | video/avi 或 video/x-msvideo |\n| MOV | .mov | url 传入视频：对象存储请设置 Content-Type 为 video/quicktime；base64 编码：请使用 video/mov，即 data:video/mov;base64, |\n| MKV | .mkv | video/x-matroska |\n\n**多模态文件大小及容量限制**\n- URL / Base64 方式传入：视频文件 ≤ 50 MB，图片文件 ≤ 10 MB，请求体 ≤ 64 MB。\n- 使用 Files API 上传：以 `mm_file://{file_id}` 形式引用，单个视频最大 512 MB。", "properties": { "type": { "type": "string", "enum": [ "base64", "url" ], "description": "媒体来源类型。" }, "media_type": { "type": "string", "description": "base64 输入时必填，例如 image/png 或 video/mp4。" }, "data": { "type": "string", "description": "base64 编码后的媒体字节。" }, "url": { "type": "string", "description": "公网 URL。视频还可使用 Files API 引用 `mm_file://{file_id}`，`file_id` 需先通过[文件上传](/api-reference/file-management-upload)接口上传视频后获取。" }, "detail": { "type": "string", "enum": [ "low", "default", "high" ], "default": "default", "description": "理解精细度，默认 default。\n\n图片输入时，单张图片粗略 token 用量估算：\n\n| detail | 粗略 token 用量 |\n| :-- | :-- |\n| low | 通常为几百 token，最高约 600 |\n| default | 通常约 1k-3k token，最高约 5k |\n| high | 通常为数千 token，最高约 15k+ |\n\n实际用量取决于图片尺寸和内容；请以 `POST /anthropic/v1/messages/count_tokens` 或响应中的 usage 为准。" }, "fps": { "type": "number", "minimum": 0.2, "maximum": 5, "default": 1, "description": "视频抽帧频率，默认 1，范围 [0.2, 5]。\n- 取值越高：对画面变化越敏感，token 花费高、速度慢。\n- 取值越低：token 花费少、速度快，但对画面变化迟钝。" }, "max_long_side_pixel": { "type": "integer", "minimum": 1, "description": "图片或视频帧最长边像素约束。" } }, "required": [ "type" ] }, "Tool": { "type": "object", "description": "Anthropic 兼容工具调用的工具定义。", "properties": { "name": { "type": "string", "description": "工具名称。" }, "description": { "type": "string", "description": "工具描述。" }, "input_schema": { "type": "object", "description": "工具输入参数 JSON Schema。" }, "cache_control": { "$ref": "#/components/schemas/CacheControl" } }, "required": [ "name", "input_schema" ] }, "ToolChoice": { "type": "object", "description": "工具选择策略。仅支持 auto 和 none。", "properties": { "type": { "type": "string", "enum": [ "auto", "none" ], "description": "工具选择类型。" } }, "required": [ "type" ] }, "CountTokensResp": { "type": "object", "description": "返回本次请求的输入 token 用量，不生成模型输出。", "properties": { "usage": { "$ref": "#/components/schemas/Usage" } } }, "ErrorResponse": { "type": "object", "description": "错误响应。统一使用 HTTP 状态码 + JSON body。流式过程中出现错误时，会以 `event: error` SSE 事件下发，body 结构与此一致；客户端应在收到 error 后停止读取并清理本次会话状态。", "properties": { "type": { "type": "string", "enum": [ "error" ], "description": "固定为 `error`。" }, "request_id": { "type": "string", "description": "本次请求的唯一标识，便于排查问题。" }, "error": { "type": "object", "properties": { "type": { "type": "string", "description": "错误类型。", "enum": [ "invalid_request_error", "authentication_error", "permission_error", "not_found_error", "request_too_large", "rate_limit_error", "api_error", "overloaded_error" ] }, "message": { "type": "string", "description": "错误详情。" } } } } } }, "securitySchemes": { "bearerAuth": { "type": "http", "scheme": "bearer", "description": "Bearer API Key 鉴权。发送 Authorization: Bearer 。如果 Authorization 和 x-api-key 同时存在，优先使用 Authorization。" }, "apiKeyAuth": { "type": "apiKey", "in": "header", "name": "x-api-key", "description": "Anthropic 兼容 API Key 鉴权。发送 x-api-key: 。推荐使用 Authorization: Bearer 。" } } } }