Cherry Studio Complete Guide: 300+ Models in One Desktop App — Local KB + MCP, Zero-Cost Unified Calling
TL;DR: Cherry Studio is an open-source desktop AI assistant (Electron, 300+ providers) that collapses OpenAI, Claude, Gemini, DeepSeek, Groq, OpenRouter into one UI, with local knowledge base, MCP tools and multi-model side-by-side comparison — bringing unified calling to the desktop.
Why Cherry Studio?
- 300+ providers: Templates for OpenAI, Anthropic, Gemini, DeepSeek, Groq, Together, OpenRouter, Ollama — paste key and go.
- Local RAG: Import PDF/Word/Markdown/Excel, vector search with citations, offline.
- MCP: Native Model Context Protocol — filesystem, web search, code execution pluggable.
- Multi-model: Send one prompt to 3-5 models in parallel, compare quality/latency.
- Local-first: SQLite + encrypted keys, no third-party relay.
Setup
Download from https://cherry-ai.com or GitHub CherryHQ/cherry-studio (Windows/macOS/Linux). Settings → Model Services → Add → pick provider → paste Key + Base URL → Check → select models. Top switcher then swaps models instantly.
OpenRouter unified entry
Type: OpenAI Compatible
Base URL: https://openrouter.ai/api/v1
Key: sk-or-v1-xxx
Models: deepseek/deepseek-r1:free, qwen/qwen3-235b:free, moonshotai/kimi-k2:free
Unified Calling
- Side-by-side: Enable multi-model, pick 3 models, send once, compare.
- Knowledge Base: Create KB → drag docs → @KB in chat → retrieval-augmented answer with citations.
- MCP: Settings → MCP → add servers (filesystem, brave-search, python) → model auto-calls tools.
vs OneAPI / LiteLLM
| Solution | Form | Layer | Best For |
|---|---|---|---|
| Cherry Studio | Desktop | Client aggregation | Personal multi-model + local RAG |
| OneAPI Gateway | Single binary | Server aggregation | Team self-host + monetize |
| LiteLLM Proxy | Python proxy | Server aggregation | Python teams, 100+ models |
| Portkey | Enterprise gateway | Server + governance | Enterprise guardrails |
Best Practices
- Pin
OpenRouter :freemodels on top for daily free use. - Alias models to short names (
r1-free,k2). - Backup
config.jsonfor migration. - Connect Ollama at
localhost:11434for offline unified UI.
Summary
- Free multi-model: Cherry + OpenRouter :free, compare 10+ models after one install.
- Local KB: Drag-and-drop RAG, no vector DB setup.
- Gateway link: Point Cherry's Base URL to OneAPI
http://localhost:3000/v1to reusemymodelrouting.
🚀 Get Started: One-Click Free API Access
Want to call all the free models above with a single API key, no need to sign up for each provider? Apishare.cc provides a unified API Key — one key, 100+ models, free models at zero cost.
👉 Register on Apishare.cc → Get your unified API Key
📊 Want to see more free model rankings? Check out the Sep 2026 Free LLM API Rankings →
About the Free API Aggregator
The models covered in this guide are all served through the APIShare free API aggregator, which gives you one key for the whole catalog.
- Full model catalog: APIShare free API directory
- Sign up for a free trial key: Register and claim your API key