Claude Compatibility Protocol

Learn how to connect clients, CLI programming tools, and Anthropic SDKs that only support Claude's native interface to OpenRely via the official Anthropic native protocol (/v1/messages).

Protocol Overview

Certain clients, command-line agents (such as Claude Code native mode, some Immersive Translate plugins, and Claude Desktop extensions) built specifically for the Anthropic ecosystem, as well as development projects using the official @anthropic-ai/sdk, typically only support the Anthropic Native Messages Protocol Specification.

OpenRely API natively supports the Anthropic /v1/messages endpoint format. Simply redirect your request Base URL to OpenRely and provide your OpenRely API Key to access world-leading models like Claude directly at high speed without network barriers.

📌 Core Connection Specifications

Parameter	Standard Value	Description
API Service Address (Base URL)	`https://api.openrely.ai`	If the client automatically appends `/v1/messages` under the hood, just enter the root domain; some tools may require `https://api.openrely.ai/v1`
Authentication Header (Headers)	`x-api-key: sk-••••`	Enter your OpenRely API Key (the system is also automatically compatible with `Authorization: Bearer sk-••••`)
Version Control Header (Version)	`anthropic-version: 2023-06-01`	Official recommended standard API version by Anthropic
Data Format	`Content-Type: application/json`	Standard JSON request and response body

⚡ Core Endpoints and Request Examples

Message Creation (Messages API)

Anthropic's official core chat message endpoint, natively supporting the Thinking deep thought process, Prompt Caching, and tool use.

Endpoint: POST /v1/messages
Full URL: https://api.openrely.ai/v1/messages

Request Example

curl -X POST "https://api.openrely.ai/v1/messages" \
  -H "x-api-key: sk-••••" \
  -H "anthropic-version: 2023-06-01" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "claude",
    "max_tokens": 1024,
    "messages": [
      { "role": "user", "content": "Hello Claude!" }
    ]
  }'

Response Example

{
  "id": "msg_01XyZ987...",
  "type": "message",
  "role": "assistant",
  "model": "claude",
  "content": [
    {
      "type": "text",
      "text": "Claude message routed."
    }
  ],
  "stop_reason": "end_turn",
  "stop_sequence": null,
  "usage": {
    "input_tokens": 11,
    "output_tokens": 18
  }
}

🛠️ Quick Client & SDK Integration

from anthropic import Anthropic

client = Anthropic(
    base_url="https://api.openrely.ai",
    api_key="sk-your-openrely-key",
)

message = client.messages.create(
    model="claude",
    max_tokens=1024,
    messages=[
        {"role": "user", "content": "Introduce yourself in one sentence"}
    ],
)

print(message.content[0].text)

import Anthropic from '@anthropic-ai/sdk';

const anthropic = new Anthropic({
  baseURL: 'https://api.openrely.ai',
  apiKey: 'sk-your-openrely-key',
});

const response = await anthropic.messages.create({
  model: 'claude',
  max_tokens: 1024,
  messages: [{ role: 'user', content: 'Hello Claude!' }],
});

console.log(response.content[0].text);

If the third-party CLI tool you use follows Anthropic's official environment variable naming conventions, you can execute the following directly in your terminal:

export ANTHROPIC_BASE_URL="https://api.openrely.ai"
export ANTHROPIC_API_KEY="sk-your-openrely-key"