DeepRoute AI for VS Code

DeepRoute AI brings local-first DeepSeek reasoning to your editor with zero configuration. It intelligently attempts to route prompts to your local Ollama instance first for fast, 100% private, and unlimited inference. If your local Ollama engine isn't active, it seamlessly switches to OpenRouter cloud APIs.

Features

• Unlimited Local AI: Automatically detects local Ollama servers (http://localhost:11434) running deepseek-r1.
• Cloud Fallback: Gracefully routes queries to OpenRouter when offline engines aren't reachable.
• Interactive Sidebar UI: Clean panel with real-time markdown streaming, code rendering, and instant copy buttons.
• Customizable Settings: Easily configure your local endpoints, cloud models, and preferred LLM engines in VS Code settings.
• Secure Key Storage: OpenRouter API keys are safely encrypted using VS Code's native SecretStorage.

Getting Started

Option 1: Local Engine (Recommended – Free & Unlimited)

1. Download and install Ollama from ollama.com.
2. Open your terminal and run: ollama run deepseek-r1:8b
3. Open the DeepRoute AI icon in your VS Code sidebar and start typing!

Option 2: Cloud API (OpenRouter)

1. Open the VS Code Command Palette (Ctrl+Shift+P / Cmd+Shift+P).
2. Run DeepRoute: Set OpenRouter API Key.
3. Enter your OpenRouter API key.

Extension Commands

Extension Settings

You can customize the following settings in your VS Code settings.json:

• deeproute.ollamaEndpoint: Base URL for your local Ollama host (default: http://localhost:11434).
• deeproute.ollamaModel: Default local Ollama model ID (default: deepseek-r1:8b).
• deeproute.cloudEndpoint: Gateway endpoint for cloud models (default: https://openrouter.ai/api/v1).
• deeproute.cloudModel: Default free cloud model ID (default: deepseek/deepseek-r1:free).

License

MIT