Claude Native Messages

In addition to the OpenAI-compatible format, YouQi AI exposes the Anthropic Claude native Messages API, so you can use the Anthropic SDK or an existing Claude integration directly — no rewrite to the OpenAI format required.

Endpoint & authentication

Item	Value
Endpoint	POST /v1/messages
Base URL	https://ai.youqi.tech
Auth	x-api-key: <YOUR_API_KEY>
Version header	anthropic-version: 2023-06-01

Difference from the OpenAI format

The Claude native API uses the x-api-key header (not Authorization: Bearer) and requires the anthropic-version header. The API key is still the sk-... key you create in the console.

Create a message

Request parameters

Parameter	Type	Required	Description
model	string	Yes	Model ID, e.g. claude-3-5-sonnet-20241022
messages	array	Yes	Conversation messages, each with role (user / assistant) and content
max_tokens	integer	Yes	Max tokens to generate (required by the Claude native API)
system	string	No	System prompt (a separate field, not part of messages)
stream	boolean	No	Whether to stream, default false
temperature	number	No	Sampling temperature, 0–1
top_p	number	No	Nucleus sampling
stop_sequences	array	No	Stop sequences
tools	array	No	Tool (function) definitions

Request example

curl https://ai.youqi.tech/v1/messages \
  -H "x-api-key: sk-YOUR_API_KEY" \
  -H "anthropic-version: 2023-06-01" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "claude-3-5-sonnet-20241022",
    "max_tokens": 1024,
    "messages": [
      { "role": "user", "content": "Introduce Hangzhou in one sentence." }
    ]
  }'

With the official Anthropic Python SDK, just point base_url to YouQi AI:

from anthropic import Anthropic

client = Anthropic(
    base_url="https://ai.youqi.tech",
    api_key="sk-YOUR_API_KEY",
)

resp = client.messages.create(
    model="claude-3-5-sonnet-20241022",
    max_tokens=1024,
    messages=[{"role": "user", "content": "Hello"}],
)
print(resp.content[0].text)

Response example

{
  "id": "msg_abc123",
  "type": "message",
  "role": "assistant",
  "model": "claude-3-5-sonnet-20241022",
  "content": [
    { "type": "text", "text": "Hangzhou is a city famed for West Lake, blending deep history with a vibrant digital economy." }
  ],
  "stop_reason": "end_turn",
  "usage": {
    "input_tokens": 18,
    "output_tokens": 22
  }
}

Streaming (SSE)

Add "stream": true to the request body and the server returns Server-Sent Events. Event types include message_start, content_block_delta, message_delta, and message_stop:

event: content_block_delta
data: {"type":"content_block_delta","index":0,"delta":{"type":"text_delta","text":"Hang"}}

Tool use

Declare callable tools in tools. When the model needs to call one, it returns a type: "tool_use" block in the response content; you execute it and send the result back as a tool_result content block with role: "user" to continue the conversation.

{
  "model": "claude-3-5-sonnet-20241022",
  "max_tokens": 1024,
  "messages": [{ "role": "user", "content": "What's the weather in Beijing today?" }],
  "tools": [
    {
      "name": "get_weather",
      "description": "Get the weather for a given city",
      "input_schema": {
        "type": "object",
        "properties": { "city": { "type": "string" } },
        "required": ["city"]
      }
    }
  ]
}

Note

If you use the OpenAI SDK, you can also call Claude models via the Chat endpoint (/v1/chat/completions) — the gateway handles protocol conversion automatically. The native Messages API suits existing Anthropic integrations or Claude-specific fields.