Cherry Studio Setup Guide

Configure a custom provider in Cherry Studio to use our API.

Download the client: cherry-ai.com (macOS, Windows, Linux)

1. Add a Custom Provider

Open Cherry Studio → Settings → Model Providers → click Add Provider:

  • Provider Name: Any name you like (e.g., "AI API Proxy")
  • Provider Type: Select OpenAI Compatible
  • API Base URL: http://137.184.200.132:3001/api/v1
  • API Key: Your API key from the dashboard
💡 If the /v1 endpoint reports an error, try http://137.184.200.132:3001 without /v1.

2. Add Models

After saving the provider, click Add Model. Cherry Studio auto-fetches models from our endpoint. Alternatively, manually add:

  • Model ID: gpt-5.6-sol — best all-around
  • Model ID: claude-sonnet-4-6 — best value (requires Endpoint Compatibility below)
  • Model ID: deepseek-v4-pro — deep reasoning
  • Model ID: qwen3.8-max — 1M context
  • Model ID: grok-4.5 — fast responses
⚠️ For Claude models: Set Endpoint Compatibility to Anthropic Compatible. This routes requests to /api/v1/messages instead of /api/v1/chat/completions. Non-Claude models use the default OpenAI Compatible setting.

3. Start Chatting

Select a model from the dropdown at the top of the chat window. Send a test message to verify everything works.

You can switch models mid-conversation — all models share the same API key and balance.

Troubleshooting

  • Connection refused: Check the Base URL. Make sure there's no trailing slash and the server is running.
  • 401 Unauthorized: Your API key may be inactive or incorrect. Verify in the dashboard.
  • Model not appearing: Click the refresh icon next to the model list. If auto-fetch fails, add models manually.
  • Empty responses: Claude models need Endpoint Compatibility set to "Anthropic Compatible". Switch it and try again.
  • Slow responses: Some models are larger and take longer. Try a faster model like gpt-5.4 or claude-haiku-4-5-20251001.