聊天

POST

/v1/messages

基础文本对话，只支持claude系列模型

支持思考模式，需要设置 thinking 参数，示例：
“thinking”:
{
"type": "enabled",
"budget_tokens": 2048
}
}

在 Claude Opus 4.7 中，自适应思考是唯一受支持的思维模式，不再接受手动传值

thinking: {type:
        "enabled", budget_tokens: N}

（强行调用将被拒绝）

请求参数

Header 参数

Body 参数application/json必填

示例

{
    "model": "claude-opus-4-6",
    "max_tokens": 1024,
    "messages": [
        {
            "role": "user",
            "content": "Hello, world"
        }
    ]
}

请求示例代码

Shell

JavaScript

Java

Swift

PHP

Python

HTTP

Objective-C

Ruby

OCaml

Dart

curl --location 'https://ai-router.bcltech.net/v1/messages' \
--header 'Authorization: {{Authorization}}' \
--header 'Content-Type: application/json' \
--data '{
    "model": "claude-opus-4-6",
    "max_tokens": 1024,
    "messages": [
        {
            "role": "user",
            "content": "Hello, world"
        }
    ]
}'

返回响应

🟢200成功

application/json

Bodyapplication/json

示例

{
    "model": "claude-opus-4-6",
    "id": "msg_018dVKipdTRa6RTXNXnW5QFa",
    "type": "message",
    "role": "assistant",
    "content": [
        {
            "type": "text",
            "text": "Hello! 👋 Welcome! How can I help you today? Whether you want to chat, ask questions, or work through something, I'm here for you. 😊"
        }
    ],
    "stop_reason": "end_turn",
    "stop_sequence": null,
    "stop_details": null,
    "usage": {
        "input_tokens": 10,
        "cache_creation_input_tokens": 0,
        "cache_read_input_tokens": 0,
        "cache_creation": {
            "ephemeral_5m_input_tokens": 0,
            "ephemeral_1h_input_tokens": 0
        },
        "output_tokens": 41,
        "service_tier": "standard",
        "inference_geo": "global"
    }
}

修改于 2026-06-30 07:40:44

Chat（聊天）

聊天（流式返回）