
Chatspeed
Any Claude Code, Any Gemini CLI
🔄 Multi-Protocol Conversion
Seamlessly convert between OpenAI, Claude, Gemini, Ollama, and more. Integrate any AI model into your preferred development environment.
🎯 Any Claude Code
Integrate free or cost-effective models (e.g., OpenRouter, Kimi, Qwen, GLM) into Claude Code, slashing development costs by over 80%.
🌐 Any Gemini CLI
Not just Claude, Gemini CLI can also integrate any model. Whether it's Claude 3 or various free/open-source models, they can all become part of your CLI workflow.
🔗 Unified MCP Entry
Aggregate all MCP (Model Context Protocol) tools into a unified proxy entry, eliminating the pain of repeatedly configuring tools across IDEs like VS Code and Cursor.
🎛️ Smart Grouping & Routing
Create different AI proxy groups and quickly switch models via endpoints (e.g., /kimi, /qwen) for flexible, efficient multi-scenario needs.
⚙️ Advanced Prompt Engineering
Each proxy group supports independent prompt injection and modification, with fine-tuned model parameters (e.g., temperature) to optimize performance in specific development environments.
🚀 Universal AI Proxy ccproxy
ccproxy
(Chat Completion Proxy
) is Chatspeed's core proxy engine, a universal adapter designed for AI development. It primarily solves two major pain points:
1. Protocol Conversion for Cost-Efficiency
It enables arbitrary protocol conversion between OpenAI-compatible formats, Gemini, Claude, Ollama, and more. This means you can seamlessly integrate any local or cloud-based free, open-source, or cost-effective model into popular development environments like Claude Code
and Gemini CLI
that require specific protocols.
✨ Special Feature: Tool Calling Compatibility Mode
For models that do not natively support tool calling (Function Calling), ccproxy
provides a compatibility mode, allowing them to understand and execute tool calling instructions, greatly expanding their capabilities.
2. Unified Entry Point for Streamlined Workflow
It aggregates all MCP tools (e.g., web browsing, file operations) installed on Chatspeed into a single, unified proxy entry point for external service. Configure once, and all applications and terminals can call them, completely eliminating repetitive setup.
You can access
ccproxy
's SSE protocol MCP viahttp://localhost:11434/sse
Key Benefits
- 🔀 Seamless Protocol Conversion: Efficiently convert between mainstream AI protocols, achieving a "universal" adaptation effect.
- 💰 Ultimate Cost Control: Flexibly use free and cost-effective models, making "using Claude Code for free" a reality.
- 🔄 Configure Once, Use Everywhere: Eliminate the tedious, repetitive configuration of MCP tools across multiple IDEs, boosting development efficiency.
- 🎯 Flexible Group Switching: Efficiently use AI-assisted programming in different scenarios through various model groups.
- 🔑 Enhanced Security: Use
ccproxy
's proxy key to effectively isolate clients from real keys, ensuring your AI account security. - 🚀 High Performance: A high-performance adapter implemented in Rust, ensuring stability and reliability.
Quick Examples
# Access the model list under the kimi group using the Claude Code protocol, {proxy-key} can be configured in the software settingscurl-H"x-api-key: {proxy-key}" http://127.0.0.1:11434/kimi/v1/models# Access the model list under the qwen group using the Gemini protocolcurl http://127.0.0.1:11434/qwen/v1beta/models?key={proxy-key}# Access the 'gemma3-12b' model (proxy alias) configured in the proxy via the Ollama protocolcurl http://localhost:11434/api/chat -d'{ "model": "gemma3-12b", "messages": [ { "role": "user", "content": "你是哪个家伙?" } ], "stream": true}'# Access the chat interface of the Ollama protocol via tool compatibility mode, where 'gemma3-12b' is the "proxy alias" configured in the softwarecurl http://localhost:11434/compat_mode/api/chat -d'{ "model": "gemma3-12b", "messages": [ { "role": "user", "content": "what is the weather in tokyo?" } ], "tools": [ { "type": "function", "function": { "name": "get_weather", "description": "Get the weather in a given city", "parameters": { "type": "object", "properties": { "city": { "type": "string", "description": "The city to get the weather for" } }, "required": ["city"] } } } ], "stream": true}'# Access the chat interface of the Dali server via the Claude protocol, with tool compatibility mode enabled# Note: The proxy key and proxy alias are configured in the software. The example here is for my local testing and cannot be run directly.curl http://localhost:11434/compat_mode/v1/messages \-H"x-api-key: cs-TosCz7A29R74yNnShYQKskxXPx9OSYd1RMQV5YzVdHvqL7ehNNoOhCVg7UTp6"\-H"Content-Type: application/json"\-H"anthropic-version: 2023-06-01"\-d'{ "model": "gemma3-12b", "messages": [{"role": "user", "content": "帮我获取下北京的天气"}], "tools": [ { "name": "get_weather", "description": "Get weather information", "input_schema": { "type": "object", "properties": { "location": {"type": "string"} } } } ], "max_tokens": 1024, "stream": true}'# Access the chat interface of the proxy server via tool compatibility mode and OpenAI compatible protocol# When testing, remember to change the model ID to the proxy alias configured in the software, and the key to your own configured key.curl http://localhost:11434/compat_mode/v1/chat/completions \-H"Authorization: Bearer cs-TosCz7A29R74yNnShYQKskxXPx9OSYd1RMQV5YzVdHvqL7ehNNoOhCVg7UTp6"\-H"Content-Type: application/json"\-d'{ "model": "gemma3-12b", "messages": [{"role": "user", "content": "帮我查下杭州有啥好玩的"}], "tools": [{ "type": "function", "function": { "name": "search_web", "description": "Search the web", "parameters": {"type": "object", "properties": {"query": {"type": "string"}}}} }], "stream": true}'# Access the chat interface of the proxy server via the Gemini protocol in tool compatibility modecurl http://localhost:11434/compat_mode/v1beta/models/gemma3-12b:streamGenerateContent?alt=sse&key=cs-TosCz7A29R74yNnShYQKskxXPx9OSYd1RMQV5YzVdHvqL7ehNNoOhCVg7UTp6 \-H"Content-Type: application/json"-d'{ "contents": [ { "role": "user", "parts": [{"text": "What is the price of GOOG stock?"}] } ], "tools": [ { "functionDeclarations": [ { "name": "get_stock_price", "description": "Get the latest price of a stock", "parameters": { "type": "OBJECT", "properties": { "symbol": { "type": "STRING", "description": "The stock ticker symbol" } }, "required": ["symbol"] } } ] } ]}'
For more
curl
examples, please refer to the ccproxy Module documentation.
🌟 More Than Just a Proxy: An All-in-One AI Assistant
Chatspeed is not only a powerful proxy platform but also a fully-featured, desktop-resident AI assistant that will transform your daily workflow.
- Smart Conversation & Creation: Engage in fluid conversations with any integrated model. Whether generating code, writing articles, or sparking creativity, it's your powerful partner.
- AI-Assisted Diagramming: Collaborate with AI to automatically generate and modify mind maps and flowcharts, helping you organize complex thoughts.
- Multi-Language Translation: Built-in powerful translation capabilities automatically detect your input language and translate it into your set target language.
- Smart Notes Knowledge Base: During AI chats, you can easily save valuable content (whether code blocks, diagrams, or text) to "Smart Notes," building your private knowledge base.
- Extensible Skill System: Beyond built-in features, you can create any new tools via the "Skills" system, continuously expanding the AI's capabilities.
Here are two example images:
📖 Documentation Structure
- Getting Started - Installation, configuration, and quick start
- ccproxy Module - Detailed ccproxy setup and usage
- MCP Integration - Model Context Protocol proxy configuration
- API Reference - Complete API documentation
🤝 Community & Support
Ready to revolutionize your AI development workflow?Get started now and experience the power of unified AI proxy management!