Twinny is a Visual Studio Code extension that provides AI-powered code completion and chat features, available for free with support for both local and API-hosted models.
The tool addresses the cost barrier of commercial AI coding assistants by offering fill-in-the-middle code completion and conversational AI features without subscription fees. It works by connecting to either a locally-hosted model via Ollama or compatible APIs, or to external providers including OpenAI, Anthropic, Mistral, Groq, and others. The extension provides real-time inline suggestions as you type, alongside a sidebar chat interface for discussing code, requesting refactoring, generating tests, and creating commit messages. Workspace embeddings enable context-aware assistance by understanding your project structure.
Twinny suits developers who want AI coding assistance without ongoing costs or who require privacy-preserving local execution. It works well for teams through its gateway feature, which allows one model server to serve multiple developers with centralized configuration and usage tracking. The extension is particularly valuable for those already using Ollama or running local language models, as it integrates seamlessly with existing setups. Unlike commercial alternatives, it offers complete offline operation and customizable API endpoints, giving users control over which models and providers they use.
Development activity shows consistent engagement with the project. The maintainers actively solicit contributions through GitHub issues and social channels. The tool receives regular updates addressing functionality and user-reported issues. The project maintains documentation including setup guides and team deployment instructions, indicating sustained attention to user onboarding and feature explanation.
