创建聊天对话(流式)

POST

/v1/chat/completions

根据对话历史创建模型响应。支持流式和非流式响应。
兼容 OpenAI Chat Completions API。

请求参数

Header 参数

Body 参数application/json必填

示例

{
    "model": "mimo-v2.5",
    "messages": [
        {
            "role": "system",
            "content": "You are a helpful assistant."
        },
        {
            "role": "user",
            "content": "Hello!你是什么模型?"
        }
    ],
    "stream": true
}

请求示例代码

Shell

JavaScript

Java

Swift

PHP

Python

HTTP

Objective-C

Ruby

OCaml

Dart

curl --location 'https://ai.tiantoken.com/v1/chat/completions' \
--header 'Accept: application/json' \
--header 'Authorization: ' \
--header 'New-Api-User: 1' \
--header 'Content-Type: application/json' \
--data '{
    "model": "mimo-v2.5",
    "messages": [
        {
            "role": "system",
            "content": "You are a helpful assistant."
        },
        {
            "role": "user",
            "content": "Hello!你是什么模型?"
        }
    ],
    "stream": true
}'

返回响应

🟢200

application/json

成功创建响应

Bodyapplication/json

示例

{
    "id": "string",
    "object": "chat.completion",
    "created": 0,
    "model": "string",
    "service_tier": "string",
    "choices": [
        {
            "index": 0,
            "message": {
                "role": "system",
                "content": "string",
                "name": "string",
                "reasoning_content": "string"
            },
            "finish_reason": "stop"
        }
    ],
    "usage": {
        "prompt_tokens": 0,
        "completion_tokens": 0,
        "total_tokens": 0,
        "prompt_tokens_details": {
            "cached_tokens": 0,
            "text_tokens": 0,
            "audio_tokens": 0,
            "image_tokens": 0
        },
        "completion_tokens_details": {
            "reasoning_tokens": 0,
            "audio_tokens": 0,
            "accepted_prediction_tokens": "string",
            "rejected_prediction_tokens": "string"
        }
    },
    "system_fingerprint": "string"
}

修改于 2026-09-11 03:45:42

Kimi 识图

创建聊天对话(非流)