Baichuan AI Platform API

https://platform.baichuan-ai.com/docs
Admin
Documentation for the Baichuan AI Platform API, detailing how to interact with its large language...
Tokens:40,840
Snippets:75
Trust Score:7.5
Update:8 months ago
Show doc for...
Tokens:
Raw
### Streaming API Example

Source: https://platform.baichuan-ai.com/docs/index

Example of how to make a streaming request to the chat completions endpoint.

```APIDOC
## POST /v1/chat/completions

### Description
This endpoint allows for chat-based completions with the option to stream the response in real-time. It's suitable for interactive applications where immediate feedback is desired.

### Method
POST

### Endpoint
/v1/chat/completions

### Parameters
#### Query Parameters
- **stream** (boolean) - Required - Set to `true` to enable streaming responses.

#### Request Body
- **model** (string) - Required - The model to use for the completion (e.g., 'Baichuan4-Turbo').
- **messages** (array) - Required - An array of message objects representing the conversation history.
  - **role** (string) - Required - The role of the message sender ('user' or 'assistant').
  - **content** (string) - Required - The content of the message.
- **temperature** (number) - Optional - Controls randomness. Lower values make the output more focused and deterministic.

### Request Example
```bash
curl -X POST https://api.baichuan-ai.com/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer ${API_KEY}" \
  -d '{
        "model": "Baichuan4-Turbo",
        "messages": [
          {
            "role": "user",
            "content": "What is the highest mountain in the world?"
          }
        ],
        "temperature": 0.3,
        "stream": true
      }'
```

### Response
#### Success Response (200)
When `stream` is `true`, the response will be a stream of Server-Sent Events (SSE). Each event will be a JSON object containing a chunk of the completion. The final event will signal the end of the stream.

- **id** (string) - Unique identifier for the completion.
- **object** (string) - Type of the object, typically 'chat.completion.chunk'.
- **created** (integer) - Timestamp of creation.
- **model** (string) - The model used for the completion.
- **choices** (array) - An array containing a single completion chunk.
  - **index** (integer) - Index of the choice (always 0 for chunks).
  - **delta** (object) - The content delta for this chunk.
    - **role** (string) - Role of the message sender ('assistant').
    - **content** (string) - The generated text content for this chunk.
  - **finish_reason** (string) - The reason the generation finished (e.g., 'stop', 'null').

#### Response Example (Streamed Events)
```
event: chat.completion.chunk
data: {"id": "chatcmpl-xxxx", "object": "chat.completion.chunk", "created": 1677652731, "model": "Baichuan4-Turbo", "choices": [{"index": 0, "delta": {"role": "assistant"}, "finish_reason": null}]}

event: chat.completion.chunk
data: {"id": "chatcmpl-xxxx", "object": "chat.completion.chunk", "created": 1677652731, "model": "Baichuan4-Turbo", "choices": [{"index": 0, "delta": {"content": "The highest mountain in the world is "}, "finish_reason": null}]}

event: chat.completion.chunk
data: {"id": "chatcmpl-xxxx", "object": "chat.completion.chunk", "created": 1677652731, "model": "Baichuan4-Turbo", "choices": [{"index": 0, "delta": {"content": "Mount Everest."}, "finish_reason": "stop"}]}

event: done
data: 

```

--------------------------------

### Function Interface Example

Source: https://platform.baichuan-ai.com/docs/index

This section illustrates an example of the function interface, which allows the AI model to interact with external tools or functions.

```APIDOC
## POST /websites/platform_baichuan-ai/function

### Description
This endpoint demonstrates the function interface, enabling the AI model to call external functions or tools. The request body specifies the function to be called and its arguments.

### Method
POST

### Endpoint
/websites/platform_baichuan-ai/function

### Parameters
#### Request Body
- **function_name** (string) - Required - The name of the function to execute.
- **arguments** (object) - Required - A JSON object containing the arguments for the function.

### Request Example
```json
{
  "function_name": "get_weather",
  "arguments": {
    "location": "Beijing",
    "unit": "celsius"
  }
}
```

### Response
#### Success Response (200)
- **result** (string) - The result returned by the executed function.

#### Response Example
```json
{
  "result": "The weather in Beijing is 25 degrees Celsius."
}
```
```

--------------------------------

### Python Native API Call Example

Source: https://platform.baichuan-ai.com/docs/index

This Python code snippet demonstrates how to make a native API call to Baichuan AI. It shows the setup for authentication and constructing the request payload, similar to the cURL example, but implemented in Python. This allows for programmatic integration into Python applications.

```python
# Placeholder for Python example code
# The actual code would involve importing a library or making HTTP requests
# using modules like 'requests'

# Example structure:
# import requests
# 
# api_key = "YOUR_API_KEY"
# url = "https://api.baichuan-ai.com/v1/chat/completions"
# headers = {
#     "Content-Type": "application/json",
#     "Authorization": f"Bearer {api_key}"
# }
# data = {
#     "model": "Baichuan4",
#     "messages": [
#         {
#             "content": "巴黎奥运会中国体育代表团共获多少金牌?",
#             "role": "user"
#         }
#     ],
#     "stream": False,
#     "tools": [
#         {
#             "type": "web_search",
#             "web_search": {
#                 "enable": True,
#                 "search_mode": "quality_first"
#             }
#         }
#     ]
# }
# 
# response = requests.post(url, headers=headers, json=data)
# print(response.json())
```

--------------------------------

### Synchronous API Request Example (Baichuan AI)

Source: https://platform.baichuan-ai.com/docs/index

This example shows how to make a synchronous POST request to the Baichuan AI chat completions endpoint. It includes setting headers for content type and authorization, and a JSON payload with messages, model selection, stream settings, and tool definitions for weather information. The `tool_choice` parameter is set to 'auto'.

```shell
curl -X POST https://api.baichuan-ai.com/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer ${API_KEY}" \
  -d '{ \
        "messages": [ \
        { "content": "北京今天天气怎么样？使用json格式输出", \
          "role": "user" \
        }], \
        "model": "Baichuan4", \
        "stream": false, \
        "tools": [ \
        { \
          "type": "function", \
          "function": { \
          "name": "get_current_weather", \
          "description": "获取当前位置天气", \
          "parameters": { \
            "type": "object", \
            "properties": { \
               "location": { \
                 "type": "string", \
                 "description": "城市或者省，如上海" \
               }, \
            "format": { \
              "type": "string", \
              "description": "要使用的温度单位。从用户位置推断。" \
            } \
            }, \
           "required": [ \
            "location", \
            "format" \
          ] \
        } \
      } \
    }, \
     { \
      "type": "function", \
      "function": { \
        "name": "get_yesterday_weather", \
        "description": "获取当前位置昨日的天气", \
        "parameters": { \
          "type": "object", \
          "properties": { \
            "location": { \
              "type": "string", \
              "description": "城市或者省，如北京" \
            }, \
            "format": { \
              "type": "string", \
              "description": "要使用的温度单位。从用户位置推断。" \
            } \
          }, \
          "required": [ \
            "location", \
            "format" \
          ] \
        } \
      } \
    } \
  ], \
  "tool_choice":"auto" \
}'
```

--------------------------------

### Synchronous API Request Example (cURL)

Source: https://platform.baichuan-ai.com/docs/index

This example demonstrates how to make a synchronous POST request to the Baichuan AI chat completions endpoint using cURL. It includes setting the content type, authorization header, and the request body with model, messages, stream, and tool configurations. The `web_search` tool is enabled for quality-first search.

```bash
curl -X POST https://api.baichuan-ai.com/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer ${API_KEY}" \
  -d '{ \
      	"model": "Baichuan4", \
      	"messages": [ \

      		{ \
      			"content": "巴黎奥运会中国体育代表团共获多少金牌?", \
      			"role": "user" \
      		} \
      	], \
      	"stream": false, \
      	"tools": [{ \
      		"type": "web_search", \
      		"web_search": { \
      			"enable": true, \
      			"search_mode": "quality_first" \
      		} \
      	}] \
      }'
```

--------------------------------

### Synchronous Interface Response Example

Source: https://platform.baichuan-ai.com/docs/index

This section provides an example of a synchronous interface response from the Baichuan AI platform, illustrating a chat completion interaction.

```APIDOC
## GET /websites/platform_baichuan-ai/sync_response

### Description
This endpoint provides an example response for a synchronous interface call, demonstrating a chat completion interaction with the Baichuan AI model.

### Method
GET

### Endpoint
/websites/platform_baichuan-ai/sync_response

### Parameters
#### Query Parameters
None

#### Request Body
None

### Request Example
None

### Response
#### Success Response (200)
- **id** (string) - Unique identifier for the chat completion.
- **object** (string) - Type of the object, typically 'chat.completion'.
- **created** (integer) - Unix timestamp indicating when the completion was created.
- **model** (string) - The name of the model used for the completion.
- **choices** (array) - A list of completion choices.
  - **index** (integer) - The index of the choice.
  - **message** (object) - The message content.
    - **role** (string) - The role of the message sender (e.g., 'assistant').
    - **content** (string) - The text content of the message.
    - **reasoning_content** (string) - Detailed reasoning behind the assistant's response.
- **usage** (object) - Information about token usage.
  - **prompt_tokens** (integer) - The number of tokens in the prompt.
  - **completion_tokens** (integer) - The number of tokens in the completion.
  - **total_tokens** (integer) - The total number of tokens used.
  - **search_count** (integer) - The number of searches performed.

#### Response Example
```json
{
    "id": "chatcmpl-ME810031AEkgoTi",
    "object": "chat.completion",
    "created": 1768027602,
    "model": "Baichuan-M3",
    "choices": [
        {
            "index": 0,
            "message": {
                "role": "assistant",
                "content": "**不建议您自行处理，而应尽快咨询儿科医生进行专业评估。** 您描述的情况（小儿感冒咳嗽、痰咳不出）在儿童中很常见，多由病毒感染（如普通感冒病毒）引起，但儿童尤其是婴幼儿的呼吸道较窄，痰液堵塞可能影响呼吸。以下是分步建议和注意事项，供您参考：\n\n💡 **关键建议与步骤：**\n1.  **年龄很重要：** 请先确认孩子的年龄。1岁以下婴儿出现咳嗽尤其需要警惕。不要给1岁以下婴儿使用蜂蜜或任何止咳药。\n2.  **基础家庭护理（安全有效）：**\n    *   **保证充足水分：** 多喝温水、凉白开、稀释的果汁或适量的口服补液盐水。水分有助于稀释痰液。观察孩子是否有脱水迹象（尿少色深、口干、精神差、眼窝凹陷、皮肤干燥无弹性）。\n    *   **保持空气湿润：** 使用加湿器（注意清洁，避免滋生细菌霉菌）或将孩子带到充满蒸汽的浴室里待几分钟（注意不要烫伤）。干燥的空气会让痰变粘稠。\n    *   **适当拍背排痰：** 您可以学习正确方法：五指并拢微弯成空心掌，由下向上、由外向内轻拍孩子的背部。饭后和睡觉时不要拍。这个方法对有一定年龄的孩子有效。\n    *   **保证休息和营养：** 充足的睡眠和均衡饮食有助于免疫系统工作。\n    *   **环境管理：** 保持室内清洁，避免已知的过敏原和刺激物（如香烟、油烟、新装修气味、宠物毛、大量灰尘）。\n    *   **关于蜂蜜：** 如果您的孩子**年龄大于1岁**，可以在睡前尝试少量（约半茶匙到1茶匙）纯蜂蜜兑温水喝，有助于缓解夜间咳嗽。**1岁以下绝对禁用蜂蜜（有肉毒杆菌中毒风险）。**\n    *   **不要随意用强效止咳药：** 尤其不要给孩子吃含有可待因或福尔可定等中枢性止咳药的成人药物。咳嗽是身体的保护机制，强行止咳可能导致痰堵在肺里引发肺炎。祛痰药（如氨溴索、乙酰半胱氨酸、愈创甘油醚等）需要根据年龄和具体情况由医生判断使用。\n    *   **观察记录：** 您可以先观察并记录以下信息，这有助于医生诊断：\n        *   孩子的具体年龄。\n        *   感冒多久了？有哪些其他症状？（比如：发烧吗？最高多少度？流什么颜色的鼻涕？精神、食欲、睡眠怎么样？）\n        *   咳嗽的情况：是干咳还是湿咳？痰是什么颜色的？（黄色、绿色可能提示有细菌感染可能，但也可能是病毒感染后期）。孩子大概多大了？（这很重要）。\n        *   有无以下需要急诊的情况：呼吸困难、喘息、嘴唇发紫、精神差、高热不退、脱水迹象、咳血、颈部僵硬或剧烈头痛、怀疑异物吸入史。\n\n📌 **补充信息与建议：**\n*   **病因考虑：** 普通感冒通常1-2周自愈。但如果咳嗽超过2-3周不好转，或者出现以下任一项，需要排查其他原因（如支气管炎、肺炎、哮喘、鼻窦炎、支原体感染、异物吸入、过敏、胃食管反流等）。\n*   **何时需要及时看医生：**\n    *   任何年龄出现呼吸困难、喘息、口唇发紫、精神差。\n    *   婴儿（尤其小于3个月）出现发热或咳嗽。\n    *   发热超过3天不退，或者高热（比如超过39℃）。\n    *   咳嗽剧烈影响睡眠或进食，或者咳出带血的痰。\n    *   有脱水迹象。\n    *   您感觉孩子病情严重。\n    *   您的孩子有基础心肺疾病、免疫抑制状态。\n    *   医生可能会：进行听诊等查体，建议做血常规、C反应蛋白、胸部X光片、病原学检查（如流感、支原体）等，并根据诊断给予合适的治疗建议。如果是细菌感染，可能需要抗生素。同时可能给予安全的对症处理建议或药物。\n*   **其他可能的处理（由医生决定）：**\n    *   退热处理：如有发热不适，可用适合儿童的对乙酰酚酚或布洛芬（注意剂量和年龄限制）。\n    *   鼻腔护理：如果同时有鼻塞流涕，用生理盐水滴鼻/喷鼻剂帮助清理鼻腔，减少鼻涕倒流刺激咽喉。\n    *   识别并发症迹象：注意是否有耳朵痛（中耳炎）、持续高热寒战、咳嗽加重伴脓痰或胸痛（肺炎）、反复发作喘息（哮喘）。\n    *   预防：注意手卫生和呼吸道卫生（咳嗽打喷嚏时用手肘遮挡），流感季考虑接种流感疫苗，保证均衡营养和适度运动增强抵抗力。\n    *   重要提示：如果您所在地区是流感流行季，或者孩子有高热、全身酸痛乏力等流感样症状，请尽快就医排查流感。流感可以引起严重并发症。\n    *   您可以做的临时缓解：让孩子多喝水，家里开加湿器。不要让孩子接触二手烟。确保孩子休息好。记录症状细节。\n\n> 举个例子：如果您告诉我“孩子3岁，感冒第5天了，一直流清鼻涕，这2天开始咳嗽，感觉喉咙有痰但咳不出来，没有发烧，精神食欲还好”，我能给您更有针对性的建议。当然，涉及隐私您不必透露所有细节，但年龄和关键症状很重要。\n\n如果方便，您可以补充更多匿名信息（年龄范围、性别、地区、大概病程和其他症状）。祝您的孩子早日康复！如有其他问题，请随时问我。🙏\n（注意：线上咨询不能替代面诊。如有任何不确定，请寻求当地医生的帮助。）",
                "reasoning_content": "嗯，用户问的是小儿感冒咳嗽，痰咳不出来怎么办。首先得确认孩子的年龄，因为不同年龄段的处理方法差异很大。比如婴幼儿和大孩子的用药和护理不一样。用户没提具体年龄，可能需要提醒他们注意年龄因素。\n\n接下来要考虑可能的原因。普通感冒通常是病毒感染，但也要排除其他情况，比如细菌感染或者过敏。特别是如果孩子有高烧、精神差、呼吸困难这些症状，可能需要及时就医。得在回答里强调危险信号。\n\n然后想到痰咳不出来的原因。孩子的咳嗽反射可能还没发育完全，尤其是小宝宝。另外痰太粘稠的话，孩子确实咳不出。这时候应该建议一些安全的处理方法，比如多喝水、保持空气湿度。蜂蜜水对1岁以上的孩子有效，但1岁以下不能用，这点必须说清楚，避免蜂蜜中毒的风险。\n\n还要考虑家庭护理措施。比如拍背排痰的手法是否正确，环境湿度是否合适。加湿器或者蒸汽浴这些方法可能有用，但要注意安全，别让孩子烫伤。另外，不要随便给孩子用止咳药，尤其是中枢性止咳药，可能掩盖症状导致痰堵住。这点很重要，得重点提醒。\n\n药物方面，得建议用户咨询医生。可能需要祛痰药，但不同年龄适用的药物不同。比如氨溴索、乙酰半胱氨酸这些，但必须根据医生判断。同时要检查是否有其他症状提示并发症，比如中耳炎、肺炎。\n\n还要想到用户可能的误区。比如用成人止咳药给孩子吃，或者过度使用抗生素。需要解释清楚抗生素对病毒性感冒无效，滥用会有耐药性。\n\n另外，观察孩子的精神状态、饮食和尿量。脱水的话症状会加重，比如尿少色深、口干、眼窝凹陷。这些细节可以帮用户判断是否需要紧急处理。\n\n最后要组织语言，既要专业又要易懂。分步骤说明：先观察症状和年龄，再基础护理，再何时就医。避免用太多术语，但关键点不能漏。比如提醒不要给1岁以下用蜂蜜，不要用含可待因的止咳药。\n\n可能还需要补充一些日常注意事项，比如避免烟雾刺激，保证休息。还有记录症状细节，比如咳嗽多久了，有没有发烧，痰的颜色，这些信息对医生诊断有帮助。\n\n总之，核心是安全有效的家庭处理，同时知道什么情况下必须看医生。不能耽误病情，也不能过度处理。\n"
            },
            "finish_reason": "stop"
        }
    ],
    "usage": {
        "prompt_tokens": 42,
        "completion_tokens": 1852,
        "total_tokens": 1894,
        "search_count": 0
    }
}
```
```

--------------------------------

### Streaming API Response Example

Source: https://platform.baichuan-ai.com/docs/index

This example demonstrates a typical streaming API response from the Baichuan platform. The response is broken down into chunks, each prefixed with 'data:'. Each chunk contains a JSON object representing a part of the assistant's reply. The stream concludes with 'data:[DONE]'.

```json
data:{"id":"chatcmpl-M633300APBknoaF","object":"chat.completion.chunk","created":1698205608,"model":"Baichuan4-Turbo","choices":[{"index":0,"delta":{"role":"assistant","content":"世界第一高峰是珠穆"},"finish_reason":""}]}

data:{"id":"chatcmpl-M633300APBknoaF","object":"chat.completion.chunk","created":1698205608,"model":"Baichuan4-Turbo","choices":[{"index":0,"delta":{"role":"assistant","content":"朗玛峰（Mount"},"finish_reason":""}]}

data:{"id":"chatcmpl-M633300APBknoaF","object":"chat.completion.chunk","created":1698205608,"model":"Baichuan4-Turbo","choices":[{"index":0,"delta":{"role":"assistant","content":" Everest），位于尼泊尔"},"finish_reason":""}]}

data:{"id":"chatcmpl-M633300APBknoaF","object":"chat.completion.chunk","created":1698205609,"model":"Baichuan4-Turbo","choices":[{"index":0,"delta":{"role":"assistant","content":"和中国边境，海拔高度"},"finish_reason":""}]}

data:{"id":"chatcmpl-M633300APBknoaF","object":"chat.completion.chunk","created":1698205609,"model":"Baichuan4-Turbo","choices":[{"index":0,"delta":{"role":"assistant","content":"为8,84"},"finish_reason":""}]}

data:{"id":"chatcmpl-M633300APBknoaF","object":"chat.completion.chunk","created":1698205609,"model":"Baichuan4-Turbo","choices":[{"index":0,"delta":{"role":"assistant","content":"8米。"},"finish_reason":"stop"}],"usage":{"prompt_tokens":6,"completion_tokens":29,"total_tokens":35,"search_count": 0}}

data:[DONE]
```

--------------------------------

### Synchronous Knowledge Base Response Example

Source: https://platform.baichuan-ai.com/docs/index

This JSON object demonstrates a synchronous response from the Baichuan AI platform's knowledge base. It includes chat completion details, usage statistics, and knowledge base citations with annotations.

```json
{
  "id": "chatcmpl-Mb65700CC3MOkhJ",
  "object": "chat.completion",
  "created": 1702351344,
  "model": "Baichuan2-Turbo",
  "choices": [
    {
      "index": 0,
      "message": {
        "role": "assistant",
        "content": "张三的毕业院校是xxx大学。"
      },
      "finish_reason": "stop"
    }
  ],
  "usage": {
    "prompt_tokens": 1278,
    "completion_tokens": 10,
    "total_tokens": 1288
  },
  "knowledge_base": {
    "cites": [
      {
        "title": "【高级Java工程师】张三.pdf",
        "content": "在职时间\n所属行业\n公司名称\n担任职位\n2018.05--至今\nxxxxxxx",
        "file_id": "file-HdcrTddtCp2Nbo50uci5rADP"
      }
    ],
    "cite_anotations": {
      "value": "张三的毕业院校是**xxx大学**^[1]^。",
      "annotations": [
        {
          "type": "file_citation",
          "text": "^[1]^",
          "start_index": 18,
          "end_index": 23,
          "file_citation": "在职时间\n所属行业\n公司名称\n担任职位"
        }
      ]
    }
  }
}
```

--------------------------------

### Knowledge Base Retrieval with Baichuan API

Source: https://platform.baichuan-ai.com/docs/index

This example demonstrates how to perform a knowledge base retrieval query using the Baichuan API. It specifies a `retrieval` tool with a knowledge base ID (`kb_ids`) and sets a low `temperature` for more deterministic responses. The query asks about the number of monsters in Journey to the West.

```python
from openai import OpenAI

client = OpenAI(
    api_key="your_api_key",
    base_url="https://api.baichuan-ai.com/v1/",
)

completion = client.chat.completions.create(
    model="Baichuan2-Turbo",
    messages=[
        {"role": "user", "content": "西游记里有多少妖怪？"}
    ],
    temperature=0.3,
    stream=True,
    extra_body={
        "tools": [{
            "type": "retrieval",
            "retrieval": {
                "kb_ids": [
                    "kb-123"
                ]
            }
        }]
    }
)

for chunk in completion:
    print(chunk.choices[0].delta)

```

--------------------------------

### Streaming API Response Example

Source: https://platform.baichuan-ai.com/docs/index

This section provides an example of a streaming API response, demonstrating how data is chunked and sent over time. It's commonly used for chat completions or other real-time text generation tasks.

```APIDOC
## GET /websites/platform_baichuan-ai/stream

### Description
This endpoint provides a streaming response, typically used for real-time text generation like chat completions. The response is divided into chunks, each prefixed with 'data:', and ends with '[DONE]'.

### Method
GET

### Endpoint
/websites/platform_baichuan-ai/stream

### Parameters
#### Query Parameters
- **prompt** (string) - Required - The input prompt for the AI model.
- **stream** (boolean) - Required - Set to true to enable streaming responses.

### Request Example
```
GET /websites/platform_baichuan-ai/stream?prompt=What is the highest mountain in the world?&stream=true
```

### Response
#### Success Response (200)
- **data** (string) - Each chunk of the streamed response, formatted as JSON.
- **[DONE]** (string) - Indicates the end of the stream.

#### Response Example
```
data:{"id":"chatcmpl-M633300APBknoaF","object":"chat.completion.chunk","created":1698205608,"model":"Baichuan4-Turbo","choices":[{"index":0,"delta":{"role":"assistant","content":"世界第一高峰是珠穆"},"finish_reason":""}]}

data:{"id":"chatcmpl-M633300APBknoaF","object":"chat.completion.chunk","created":1698205608,"model":"Baichuan4-Turbo","choices":[{"index":0,"delta":{"role":"assistant","content":"朗玛峰（Mount"},"finish_reason":""}]}

data:[DONE]
```
```

--------------------------------

### Synchronous Chat Completion Request Example (Baichuan API)

Source: https://platform.baichuan-ai.com/docs/index

Demonstrates a synchronous POST request to the Baichuan AI chat completions endpoint. It includes setting headers for content type and authorization, and a JSON payload with messages, model specification, stream setting, tool definitions, and tool choice.

```curl
curl -X POST https://api.baichuan-ai.com/v1/chat/completions \
-H "Content-Type: application/json" \
-H "Authorization: Bearer ${API_KEY}" \
-d '{
  "messages": [
    {
      "role": "user",
      "content": "北京今天天气怎么样？使用json格式输出"
    },{
      "role": "assistant",
      "tool_calls": [
        {
          "id": "ABCF01AEBHpopWh",
          "type": "function",
          "function": {
            "name": "get_current_weather",
            "arguments": "{\"location\": \"北京\", \"format\": \"json\"}"
          }
        }
      ]
    },{
      "role": "tool",
      "content": "多云转晴，风力1级，温度：最高20度，最低12度",
      "tool_call_id": "ABCF01AEBHpopWh"
    }
  ],
  "model": "Baichuan4",
  "stream": false,
  "tools": [
    {
      "type": "function",
      "function": {
        "name": "get_current_weather",
        "description": "获取当前位置天气",
        "parameters": {
          "type": "object",
          "properties": {
            "location": {
              "type": "string",
              "description": "城市或者省，如上海"
            },
            "format": {
              "type": "string",
              "description": "要使用的温度单位。从用户位置推断。"
            }
          },
          "required": [
            "location",
            "format"
          ]
        }
      }
    },
    {
      "type": "function",
      "function": {
        "name": "get_yesterday_weather",
        "description": "获取当前位置昨日的天气",
        "parameters": {
          "type": "object",
          "properties": {
            "location": {
              "type": "string",
              "description": "城市或者省，如北京"
            },
            "format": {
              "type": "string",
              "description": "要使用的温度单位。从用户位置推断。"
            }
          },
          "required": [
            "location",
            "format"
          ]
        }
      }
    }
  ],
  "tool_choice":"auto"
}'
```

--------------------------------

### Synchronous API Request Example (cURL)

Source: https://platform.baichuan-ai.com/docs/index

This snippet demonstrates how to make a synchronous API request to the Baichuan AI chat completions endpoint using cURL. It includes the necessary headers and a JSON payload with the model and messages.

```bash
curl -X POST https://api.baichuan-ai.com/v1/chat/completions\
     -H "Content-Type: application/json"\
     -H "Authorization: Bearer ${API_KEY}"\
     -d '{
          "model": "Baichuan-M3",
          "messages": [
            {
               "content": "小儿感冒咳嗽，痰咳不出来怎么办？",
               "role": "user"
            }
          ],
          "stream": false
        }'
```

--------------------------------

### Streaming Chat Completion Request Example (cURL)

Source: https://platform.baichuan-ai.com/docs/index

This cURL command demonstrates how to make a streaming request to the Baichuan AI chat completions endpoint. It includes the model, messages, temperature, and the crucial 'stream: true' parameter to enable streaming responses.

```curl
curl -X POST https://api.baichuan-ai.com/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer ${API_KEY}" \
  -d '{
        "model": "Baichuan4-Turbo",
        "messages": [
          {
            "role": "user",
            "content": "世界第一高峰是?"
          }
        ],
        "temperature": 0.3,
        "stream": true
      }'
```

--------------------------------

### Synchronous API Response Example (Baichuan AI)

Source: https://platform.baichuan-ai.com/docs/index

This example illustrates a typical JSON response from the Baichuan AI chat completions endpoint for a synchronous request. It includes response metadata like ID, object type, creation timestamp, and model used. The 'choices' array contains the assistant's message, which in this case includes 'tool_calls' indicating that the model has identified a function to call based on the input and defined tools. The 'finish_reason' is 'tool_calls'.

```json
{ \
  "id": "chatcmpl-Mf89d015KICDojA", \
  "object": "chat.completion", \
  "created": 1716199933, \
  "model": "Baichuan4", \
  "choices": [ \
    { \
      "index": 0, \
      "message": { \
        "role": "assistant", \
        "tool_calls": [ \
          { \
            "id": "71f7015KICDoskJ", \
            "type": "function", \
            "function": { \
              "name": "get_current_weather", \
              "arguments": "{\"format\": \"json\", \"location\": \"北京\"}" \
            } \
          } \
        ] \
      }, \
      "finish_reason": "tool_calls" \
    } \
  ], \
  "usage": { \
    "prompt_tokens": 136, \
    "completion_tokens": 22, \
    "total_tokens": 158 \
  } \
}
```

--------------------------------

### Synchronous Chat Completion Response Example (Baichuan API)

Source: https://platform.baichuan-ai.com/docs/index

Illustrates the JSON response structure received from the Baichuan AI chat completions endpoint. It includes fields for the response ID, object type, creation timestamp, model used, choices (containing the assistant's message and finish reason), and token usage statistics.

```json
{
  "id": "chatcmpl-M5A4C01AEBNZoLr",
  "object": "chat.completion",
  "created": 1728876215,
  "model": "Baichuan4",
  "choices": [
    {
      "index": 0,
      "message": {
        "role": "assistant",
        "content": "{\"天气\": \"多云转晴\", \"风力\": \"1级\", \"温度\": {\"最高\": \"20度\", \"最低\": \"12度\"}}"
      },
      "finish_reason": "stop"
    }
  ],
  "usage": {
    "prompt_tokens": 254,
    "completion_tokens": 35,
    "total_tokens": 289
  }
}
```

--------------------------------

### Get Parsed File Content Response Example

Source: https://platform.baichuan-ai.com/docs/file

This JSON object represents the response from the file parsing content retrieval API. It indicates the status of the parsing process (e.g., 'online' for success) and provides the extracted 'content' from the document.

```json
{
    "status": "online",
    "content": "xxx"
}
```

--------------------------------

### Get Parsed File Content Request Example (cURL)

Source: https://platform.baichuan-ai.com/docs/file

This cURL command demonstrates how to request the parsed content of a specific file from the Baichuan AI API. It includes the file ID in the URL and the necessary Authorization header with your API key.

```bash
curl https://api.baichuan-ai.com/v1/files/file-abc123/parsed-content \
  -H "Authorization: Bearer $API_KEY"
```

--------------------------------

### Web Search with Baichuan AI using Python SDK

Source: https://platform.baichuan-ai.com/docs/index

This Python code snippet demonstrates how to perform a web search using the Baichuan AI platform's API via the OpenAI SDK. It configures the client with an API key and base URL, then makes a chat completion request specifying the 'web_search' tool. The output is streamed and printed chunk by chunk.

```python
from openai import OpenAI

client = OpenAI(
    api_key="your_api_key",
    base_url="https://api.baichuan-ai.com/v1/",
)

completion = client.chat.completions.create(
    model="Baichuan4",
    messages=[
        {"role": "user", "content": "北京今天天气怎么样"}
    ],
    temperature=0.3,
    stream=True,
    extra_body={
        "tools": [{
            "type": "web_search",
            "web_search": {
                "enable": True,
                "search_mode": "performance_first"
            }
        }]
    }

)

for chunk in completion:
    print(chunk.choices[0].delta)

```
Tokens:
Raw