🤖 A unified command-line interface for multiple AI providers including Gemini, OpenAI, Anthropic, Doubao, and more.
- 🔄 Multi-Provider Support: Gemini, OpenAI, Anthropic, Doubao, DeepSeek, and any OpenAI-compatible API
- 💭 Thinking Mode: Real-time display of AI reasoning process (when supported)
- 🔧 Universal Configuration: Simple environment variables work across all providers
- 🚀 Streaming Responses: Real-time response streaming with thinking process visualization
- 📝 Context Preservation: Maintains conversation history across interactions
- 🛠️ Tool Integration: Supports function calling and tool usage
- 🎯 Easy Setup: Get started in seconds with any AI provider
# Global installation
npm install -g universal-ai-cli
# Or use directly with npx
npx universal-ai-cli# Start interactive mode
uai
# Or use the full command
universal-aiCreate a .env file or set environment variables:
# Example: Using Gemini
AI_PROVIDER=gemini
AI_API_KEY=your_gemini_api_key
AI_MODEL=gemini-2.5-pro
# Example: Using OpenAI
AI_PROVIDER=openai
AI_API_KEY=your_openai_api_key
AI_MODEL=gpt-4o
AI_BASE_URL=https://api.openai.com/v1
# Example: Using Doubao
AI_PROVIDER=doubao
AI_API_KEY=your_doubao_api_key
AI_MODEL=doubao-pro-4k
AI_BASE_URL=https://ark.cn-beijing.volces.com/api/v3
# Example: Using Anthropic
AI_PROVIDER=anthropic
AI_API_KEY=your_anthropic_api_key
AI_MODEL=claude-3-5-sonnet-20241022
AI_BASE_URL=https://api.anthropic.com| Provider | Status | Thinking Mode | Models |
|---|---|---|---|
| Gemini | ✅ Full Support | ✅ Native | gemini-2.5-pro, gemini-2.5-flash |
| OpenAI | ✅ Full Support | ✅ o1 Series | gpt-4o, gpt-4, o1-preview, o1-mini |
| Anthropic | ✅ Full Support | ❌ | claude-3-5-sonnet, claude-3-haiku |
| Doubao | ✅ Full Support | ✅ Thinking Models | doubao-pro-4k, doubao-thinking |
| DeepSeek | ✅ OpenAI Compatible | ✅ V3 Series | deepseek-chat, deepseek-coder |
| Custom | ✅ OpenAI Compatible | Any OpenAI-compatible API |
| Variable | Required | Description | Example |
|---|---|---|---|
AI_PROVIDER |
✅ | AI service provider | gemini, openai, anthropic, doubao |
AI_API_KEY |
✅ | API key for the provider | your_api_key_here |
AI_MODEL |
✅ | Model name | gemini-2.5-pro, gpt-4o |
AI_BASE_URL |
❌ | Custom API endpoint | https://api.example.com/v1 |
AI_THINK_SUPPORT |
❌ | Force enable/disable thinking | true, false |
AI_EMBEDDING_MODEL |
❌ | Custom embedding model | text-embedding-3-small |
When enabled, you'll see the AI's reasoning process in real-time:
🤔 Thinking: Let me analyze this step by step...
1. First, I need to understand the user's request
2. Then I'll search for relevant information
3. Finally, I'll formulate a comprehensive response
💬 Response: Based on my analysis, here's what I found...
# Using a custom OpenAI-compatible API
AI_PROVIDER=openai
AI_API_KEY=your_custom_key
AI_MODEL=your_custom_model
AI_BASE_URL=https://your-api-endpoint.com/v1# Enable debug logging
DEBUG=1 uai
# Or with environment variable
export DEBUG=1
uaiWe welcome contributions! Please see our Contributing Guide for details.
MIT License - see LICENSE file for details.
Based on the excellent Gemini CLI by Google, extended to support multiple AI providers.