OpenClaw

Overview

OpenClaw is an open-source personal AI assistant platform that connects to 25+ messaging channels (Telegram, Discord, Slack, WeChat, and more). It supports OpenAI-compatible providers, so you can wire it to Yunxin's POST /v1/chat/completions endpoint.

GitHub: https://github.com/nicepkg/openclaw

Prerequisites

• Node.js 18+ installed
• A Yunxin API key from Dashboard → API Keys (keys begin with sk-)
• A model id that supports chat completions — discover ids via the Models API

Installation

npm install -g openclaw

Or run without installing:

npx openclaw

Configuration

OpenClaw is configured via ~/.openclaw/openclaw.json. Add Yunxin as an OpenAI-type model provider, using the OpenAI base URL (with the /v1 suffix) and your sk- key:

{
  "modelProviders": {
    "yunxin": {
      "type": "openai",
      "apiKey": "sk-your-yunxin-key",
      "baseURL": "https://api.yuhuanstudio.com/v1"
    }
  },
  "defaultModelProvider": "yunxin",
  "defaultModel": "model-id"
}

Multiple models

Configure different models for different use cases:

{
  "modelProviders": {
    "yunxin": {
      "type": "openai",
      "apiKey": "sk-your-yunxin-key",
      "baseURL": "https://api.yuhuanstudio.com/v1"
    }
  },
  "defaultModelProvider": "yunxin",
  "models": {
    "default": "model-id",
    "coding": "coding-model-id",
    "creative": "creative-model-id"
  }
}

Connecting messaging channels

After configuring Yunxin as your model provider, enable your preferred channels.

Telegram

{
  "channels": {
    "telegram": {
      "enabled": true,
      "botToken": "YOUR_TELEGRAM_BOT_TOKEN"
    }
  }
}

Discord

{
  "channels": {
    "discord": {
      "enabled": true,
      "botToken": "YOUR_DISCORD_BOT_TOKEN"
    }
  }
}

Usage

# Start the service
openclaw start

# One-off message
openclaw chat "Hello, how are you?"

# Interactive mode
openclaw chat

Supported features

When OpenClaw runs against Yunxin's Chat Completions endpoint:

Feature availability depends on the chosen model — check each model's capabilities via the Models API.

Recommended configuration

1. Use a fast model for real-time channels to minimize response latency.
2. Set clear system prompts to define the assistant's persona.
3. Mind your rate limit. High-traffic channels can hit your tier's per-minute request limit — watch X-RateLimit-Remaining and pace accordingly.

Troubleshooting

Connection errors

1. Verify the apiKey in openclaw.json is a valid sk- key.
2. Confirm baseURL is https://api.yuhuanstudio.com/v1 (with /v1, since this is the OpenAI-format provider).
3. Confirm your account has sufficient credit balance — 402 insufficient_balance means top up.

Rate limiting

Yunxin's per-minute request limit depends on your tier (Free 10, Basic 50, Fellow/Pro 100, Enterprise custom). On 429, honor the Retry-After header. See Rate Limits.

Resources

• OpenClaw GitHub
• Chat Completions
• Authentication
• Rate Limits