# 通用对话 ## OpenAPI Specification ```yaml openapi: 3.0.1 info: title: '' description: '' version: 1.0.0 paths: /v1/chat/completions: post: summary: 通用对话 deprecated: false description: 详细说明参考：[API接口文档](https://docs.link-ai.tech/platform/api/chat) tags: - 智能体对话 parameters: - name: Authorization in: header description: >- 在 [LinkAI平台-应用接入](https://link-ai.tech/console/interface) 生成API Key并填入 required: true example: Bearer YOUR_API_KEY schema: type: string requestBody: content: application/json: schema: type: object properties: app_code: type: string title: 应用code description: 应用、工作流或超级AI助理的 code。若不填则表示不绑定具体应用，将请求直接传递给模型 messages: type: array items: type: object properties: role: type: string title: 角色 description: 支持填写 system, user, assistant 中的一个，分别表示系统提示词、用户问题、AI回复 content: type: string title: 内容 description: 文本内容 x-apifox-orders: - role - content required: - role - content title: 消息列表 description: 消息上下文列表，可以传入历史对话记忆，每一轮包含一条user消息和一条assistant消息，最后一条为用户的问题 model: type: string title: 模型编码 description: >- 不传则使用应用的默认模型，所有支持模型见 [模型列表](https://link-ai.tech/console/models) top_p: type: integer title: 模型采样范围 description: 控制模型采样范围，默认值为 1 temperature: type: number title: 温度 description: 默认为应用中配置的温度。温度越高回复越具有创意和不确定性，温度越低则回复更严谨 frequency_penalty: type: number title: 频率惩罚项 description: 该参数越大则更倾向于产生不同的内容，范围为 [-2, 2]，默认值为 0 presence_penalty: type: number title: 存在惩罚项 description: 该参数越大则更倾向于产生不同的内容，范围为 [-2, 2]，默认值为 0 stream: type: boolean title: 是否流式输出 description: 默认值为 false stream_options: type: object properties: include_usage: type: boolean description: 设置为true时流式输出最后一个chunk将包含 token 用量信息 title: 流式Token用量信息 x-apifox-orders: - include_usage title: 流式输出选项 description: 流式输出相关选项 tool_choice: type: string description: auto:自动选择, none:不会调用, required:必须调用一个或多个tool title: 工具选择模式 tools: type: array items: type: object properties: {} x-apifox-orders: [] title: 工具列表 description: 需要模型调用的工具列表 response_format: type: object properties: type: type: string title: 格式类型 description: 'json_object: 以JSON格式输出，text: 文本格式输出' x-apifox-orders: - type title: 输出格式 description: 设置模型的输出格式 thinking: type: object properties: type: type: string title: 思考开关 description: enabled:开启, disabled:关闭 x-apifox-orders: - type title: 深度思考 reasoning_effort: type: string description: >- 推理强度，控制推理模型在思考阶段消耗的 token 数。常用取值 minimal / low / medium / high。仅对支持该参数的推理模型生效 title: 推理强度 x-apifox-orders: - messages - app_code - model - temperature - top_p - frequency_penalty - presence_penalty - stream - stream_options - tool_choice - tools - response_format - thinking - reasoning_effort required: - messages example: app_code: '' messages: - role: user content: 你好 responses: '200': description: '' content: application/json: schema: type: object properties: choices: type: array items: type: object properties: index: type: integer message: type: object properties: role: type: string title: 角色 description: 响应中该字段值为 assistant content: type: string title: 响应内容 description: AI回复结果 required: - role - content x-apifox-orders: - role - content x-apifox-orders: - index - message usage: type: object properties: prompt_tokens: type: integer title: 请求token数 completion_tokens: type: integer title: 响应token数 total_tokens: type: integer title: 总token数 description: >- 总token计算包含 请求 和 响应 中的总token数， 其中请求又包含 应用设定、历史对话、知识库内容、用户问题 required: - prompt_tokens - completion_tokens - total_tokens x-apifox-orders: - prompt_tokens - completion_tokens - total_tokens 01J1H26ATPJJ2T82RT7TC6DVMQ: type: string required: - choices - usage - 01J1H26ATPJJ2T82RT7TC6DVMQ x-apifox-orders: - 01J1H26ATPJJ2T82RT7TC6DVMQ - choices - usage example: choices: - message: role: assistant content: 你好啊！有什么我可以帮助你的吗？ usage: prompt_tokens: 32 completion_tokens: 24 total_tokens: 56 headers: {} x-apifox-name: 成功 '400': description: '' content: application/json: schema: type: object properties: error: type: object properties: type: type: string message: type: string x-apifox-orders: - type - message required: - error x-apifox-orders: - error example: error: type: invalid_request_error message: 'Invalid request: user message content is empty' headers: {} x-apifox-name: 参数不正确 '401': description: '' content: application/json: schema: type: object properties: error: type: object properties: type: type: string message: type: string x-apifox-orders: - type - message required: - error x-apifox-orders: - error example: error: type: invalid_request_error message: >- Authentication failed, please check your API_KEY in https://link-ai.tech/console/interface headers: {} x-apifox-name: API Key鉴权失败 '402': description: '' content: application/json: schema: type: object properties: error: type: object properties: message: type: string required: - message x-apifox-orders: - message x-apifox-orders: - error example: error: message: App not found, please check your app_code param headers: {} x-apifox-name: 应用不存在 '403': description: '' content: application/json: schema: type: object properties: error: type: object properties: message: type: string required: - message x-apifox-orders: - message x-apifox-orders: - error example: error: message: No app access permissions headers: {} x-apifox-name: 应用无访问权限 '406': description: '' content: application/json: schema: type: object properties: error: type: object properties: message: type: string required: - message x-apifox-orders: - message x-apifox-orders: - error example: error: message: >- You exceeded your current quota, please check your balance in https://link-ai.tech/console/account headers: {} x-apifox-name: 余额不足 '409': description: '' content: application/json: schema: type: object properties: error: type: object properties: message: type: string required: - message x-apifox-orders: - message x-apifox-orders: - error example: error: message: Content sensitivity detection failed (内容合规检查不通过) headers: {} x-apifox-name: 内容审核不通过 '503': description: '' content: application/json: schema: type: object properties: error: type: object properties: message: type: string required: - message x-apifox-orders: - message x-apifox-orders: - error example: error: message: >- Server is too busy, please retry or contact the customer service headers: {} x-apifox-name: 服务异常 x-200:成功-流式输出: description: '' content: text/event-stream: schema: type: object properties: {} example: >- data: {"choices": [{"index": 0, "delta": {"content": "你好！"}, "finish_reason": null}], "session_id": null} data: {"choices": [{"index": 0, "delta": {"content": "我能"}, "finish_reason": null}], "session_id": null} data: {"choices": [{"index": 0, "delta": {"content": "为你"}, "finish_reason": null}], "session_id": null} data: {"choices": [{"index": 0, "delta": {"content": "做些什么？"}, "finish_reason": null}], "session_id": null} data: {"choices": [{"index": 0, "delta": {}, "finish_reason": "stop", "usage": {"prompt_tokens": 9, "completion_tokens": 6, "total_tokens": 15}}], "session_id": null} data: [DONE] headers: {} x-apifox-name: 成功-流式输出 security: - bearerAuth: [] x-apifox: schemeGroups: - id: fxalgFNckA8_wiEiIOIkk schemeIds: - bearerAuth required: true use: id: fxalgFNckA8_wiEiIOIkk scopes: fxalgFNckA8_wiEiIOIkk: bearerAuth: [] x-apifox-folder: 智能体对话 x-apifox-status: released x-run-in-apifox: https://app.apifox.com/web/project/3294133/apis/api-110597681-run components: schemas: {} securitySchemes: bearerAuth: type: bearer scheme: bearer description: LinkAI API Key，可在 https://link-ai.tech/console/interface 创建 servers: - url: https://api.link-ai.tech description: 正式环境 security: - bearerAuth: [] x-apifox: schemeGroups: - id: fxalgFNckA8_wiEiIOIkk schemeIds: - bearerAuth required: true use: id: fxalgFNckA8_wiEiIOIkk scopes: fxalgFNckA8_wiEiIOIkk: bearerAuth: [] ```