API 开发文档

使用概述

LLM-API提供兼容 OpenAI 标准的 API 接口，支持 Chat Completions、流式输出、工具调用、结构化输出等完整功能。您可以使用 OpenAI SDK 或任何兼容客户端无缝接入。

快速接入

Python 快速示例

from openai import OpenAI

client = OpenAI(
    api_key="your-api-key",
    base_url="https://api.silom.cn/llm-api/v1"
)

response = client.chat.completions.create(
    model="gpt-5.4",  # 或 claude-sonnet-4-6, gemini-3.5-flash 等
    messages=[
        {"role": "system", "content": "You are a helpful assistant."},
        {"role": "user", "content": "你好，介绍一下自己"}
    ]
)

print(response.choices[0].message.content)

cURL 快速示例

curl https://api.silom.cn/llm-api/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer your-api-key" \
  -d '{
    "model": "gpt-5.4",
    "messages": [
      {"role": "user", "content": "Hello!"}
    ]
  }'

Base URL 地址

根据您使用的客户端协议，选择对应的 Base URL：

认证方式

所有 API 请求需在 HTTP Header 中携带 API Key：

Authorization: Bearer sk-xxxxxxxxxxxxxxxxxxxxxxxx

可用模型列表

支持 OpenAI、Anthropic Claude、Google Gemini、xAI Grok、字节跳动等 17+ 模型，按量计费。

OpenAI

Anthropic Claude

Google Gemini

xAI Grok

字节跳动 Doubao

获取模型列表 API

通过 API 动态获取当前可用的模型列表：

curl https://api.silom.cn/llm-api/v1/models \
  -H "Authorization: Bearer your-api-key"

响应示例

{
  "object": "list",
  "data": [
    {
      "id": "gpt-5.4",
      "object": "model",
      "owned_by": "openai"
    },
    {
      "id": "claude-sonnet-4-6",
      "object": "model",
      "owned_by": "anthropic"
    }
  ]
}

Chat Completions API

核心对话接口，兼容 OpenAI /v1/chat/completions 标准。

请求参数

Python 示例

from openai import OpenAI

client = OpenAI(
    api_key="your-api-key",
    base_url="https://api.silom.cn/llm-api/v1"
)

response = client.chat.completions.create(
    model="gpt-5.4",
    messages=[
        {"role": "system", "content": "You are a helpful assistant."},
        {"role": "user", "content": "解释什么是 REST API"}
    ],
    temperature=0.7,
    max_tokens=1000
)

print(response.choices[0].message.content)
print(f"Tokens: {response.usage.total_tokens}")

cURL 示例

curl https://api.silom.cn/llm-api/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer your-api-key" \
  -d '{
    "model": "gpt-5.4",
    "messages": [
      {"role": "system", "content": "You are a helpful assistant."},
      {"role": "user", "content": "解释什么是 REST API"}
    ],
    "temperature": 0.7,
    "max_tokens": 1000
  }'

响应格式

{
  "id": "chatcmpl-xxx",
  "object": "chat.completion",
  "created": 1700000000,
  "model": "gpt-5.4",
  "choices": [
    {
      "index": 0,
      "message": {
        "role": "assistant",
        "content": "REST API 是一种..."
      },
      "finish_reason": "stop"
    }
  ],
  "usage": {
    "prompt_tokens": 15,
    "completion_tokens": 120,
    "total_tokens": 135
  }
}

流式输出 (Streaming)

设置 stream: true 启用 SSE 流式返回，适用于实时对话场景。

Python 流式示例

from openai import OpenAI

client = OpenAI(
    api_key="your-api-key",
    base_url="https://api.silom.cn/llm-api/v1"
)

stream = client.chat.completions.create(
    model="gpt-5.4",
    messages=[{"role": "user", "content": "写一首关于春天的诗"}],
    stream=True
)

for chunk in stream:
    if chunk.choices[0].delta.content:
        print(chunk.choices[0].delta.content, end="", flush=True)

cURL 流式示例

curl https://api.silom.cn/llm-api/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer your-api-key" \
  -d '{
    "model": "gpt-5.4",
    "messages": [{"role": "user", "content": "写一首关于春天的诗"}],
    "stream": true
  }'

流式响应格式

data: {"id":"chatcmpl-xxx","choices":[{"delta":{"content":"春"},"index":0}]}

data: {"id":"chatcmpl-xxx","choices":[{"delta":{"content":"风"},"index":0}]}

data: {"id":"chatcmpl-xxx","choices":[{"delta":{},"finish_reason":"stop","index":0}]}

data: [DONE]

结构化输出 (Structured Output)

通过 response_format 参数强制模型输出符合 JSON Schema 的结构化数据。

使用方式

from openai import OpenAI

client = OpenAI(
    api_key="your-api-key",
    base_url="https://api.silom.cn/llm-api/v1"
)

response = client.chat.completions.create(
    model="gpt-5.4",
    messages=[
        {"role": "user", "content": "提取以下文本中的人名和地点：张三在北京参加了会议"}
    ],
    response_format={
        "type": "json_schema",
        "json_schema": {
            "name": "extraction",
            "schema": {
                "type": "object",
                "properties": {
                    "names": {
                        "type": "array",
                        "items": {"type": "string"}
                    },
                    "locations": {
                        "type": "array",
                        "items": {"type": "string"}
                    }
                },
                "required": ["names", "locations"]
            }
        }
    }
)

print(response.choices[0].message.content)
# {"names": ["张三"], "locations": ["北京"]}

工具调用 (Function Calling)

让模型调用外部工具/函数，实现 AI 与业务逻辑的集成。

定义工具

from openai import OpenAI

client = OpenAI(
    api_key="your-api-key",
    base_url="https://api.silom.cn/llm-api/v1"
)

tools = [
    {
        "type": "function",
        "function": {
            "name": "get_weather",
            "description": "获取指定城市的天气信息",
            "parameters": {
                "type": "object",
                "properties": {
                    "city": {
                        "type": "string",
                        "description": "城市名称，如：北京"
                    }
                },
                "required": ["city"]
            }
        }
    }
]

response = client.chat.completions.create(
    model="gpt-5.4",
    messages=[{"role": "user", "content": "北京今天天气怎么样？"}],
    tools=tools,
    tool_choice="auto"
)

# 模型返回工具调用请求
tool_call = response.choices[0].message.tool_calls[0]
print(tool_call.function.name)      # "get_weather"
print(tool_call.function.arguments) # {"city": "北京"}

返回工具结果

# 执行工具后，将结果返回给模型
response2 = client.chat.completions.create(
    model="gpt-5.4",
    messages=[
        {"role": "user", "content": "北京今天天气怎么样？"},
        response.choices[0].message,  # 包含 tool_calls
        {
            "role": "tool",
            "tool_call_id": tool_call.id,
            "content": '{"temperature": "25°C", "condition": "晴"}'
        }
    ],
    tools=tools
)

print(response2.choices[0].message.content)
# "北京今天天气晴朗，气温 25°C..."

Responses API

Responses API 是 OpenAI 推出的新一代对话接口，提供更简洁的请求/响应格式。我们兼容此接口。

请求示例

curl https://api.silom.cn/llm-api/v1/responses \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer your-api-key" \
  -d '{
    "model": "gpt-5.4",
    "input": "用一句话解释量子计算"
  }'

多轮对话

curl https://api.silom.cn/llm-api/v1/responses \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer your-api-key" \
  -d '{
    "model": "gpt-5.4",
    "input": [
      {"role": "user", "content": "什么是机器学习？"},
      {"role": "assistant", "content": "机器学习是..."},
      {"role": "user", "content": "能举个例子吗？"}
    ]
  }'

Claude Code 使用指南

Claude Code 使用 Anthropic 协议，Base URL 不带 /v1。推荐使用 CC Switch 工具一键配置。

方式一：CC Switch（推荐）

方式二：手动配置环境变量

# macOS / Linux
export ANTHROPIC_BASE_URL="https://api.silom.cn/llm-api"
export ANTHROPIC_API_KEY="sk-your-api-key"

# Windows PowerShell
$env:ANTHROPIC_BASE_URL = "https://api.silom.cn/llm-api"
$env:ANTHROPIC_API_KEY = "sk-your-api-key"

支持的 Claude 模型

详细配置教程见 Claude Code 专属接入页

Codex 使用指南

OpenAI Codex CLI 使用 OpenAI 兼容协议，Base URL 带 /v1。

方式一：CC Switch

方式二：手动配置

export OPENAI_BASE_URL="https://api.silom.cn/llm-api/v1"
export OPENAI_API_KEY="sk-your-api-key"

详细配置教程见 Codex 专属接入页

QClaw 使用指南

QClaw 是一款 AI 编程助手，支持通过 OpenAI 兼容接口接入。

配置步骤

个人使用流程及规则

代理人使用规则

代理人（子代理）可通过邀请链接发展用户，获取佣金分成。

错误码

常见问题