Overview: Void is an open-source fork of Visual Studio Code designed to integrate AI agents directly into the coding workflow while prioritizing data privacy. Unlike proprietary tools that send code to third-party servers, Void sends messages directly to AI providers without retaining user data. It enables developers to use models from OpenAI, Claude, or host their own locally without vendor lock-in. Built on the VS Code codebase, Void offers a familiar interface with enhanced AI capabilities for real-time code assistance, change visualization, and checkpointing. This project is ideal for developers who value open-source software, data sovereignty, and want to experiment with AI-assisted coding without compromising privacy.
Due to a strategic pause in development, Void is no longer actively maintained. While existing installations will continue functioning, features may degrade over time due to lack of updates. The team encourages users to build their own customized versions using the provided guides and tooling, focusing on innovation rather than feature parity.
What You Get
- Direct LLM integration without data retention - Void sends AI queries directly to providers like OpenAI or Claude and does not store your code or prompts, ensuring privacy-sensitive workflows.
- Local model hosting support - You can run and connect any LLM locally (e.g., via Ollama or vLLM) without relying on cloud APIs, giving full control over model selection and inference.
- AI-powered code assistance - Leverage AI agents within the editor for real-time suggestions, code generation, and refactoring, similar to Cursor or GitHub Copilot.
- Change checkpointing and visualization - Visualize and revert code changes through a built-in checkpoint system, helping track AI-assisted edits and experiment safely.
- VS Code compatibility - Built on the official VS Code codebase, allowing use of existing extensions, themes, and keybindings with minimal disruption.
Common Use Cases
- Building privacy-conscious AI-assisted applications - Developers in regulated industries (healthcare, finance) use Void to integrate LLMs into their workflow without sending proprietary code to external servers.
- Experimenting with local LLMs in development - Engineers testing models like Llama 3 or Mistral via Ollama use Void to evaluate AI coding performance without cloud dependency.
- Problem → Solution flow: ‘I don’t trust AI tools that store my code’ → ‘Use Void to get Copilot-like features with zero data retention’ - Users concerned about code leakage in SaaS AI tools switch to Void for secure, on-device or self-hosted AI assistance.
- Team workflow: Research teams building custom AI editors - Developers fork Void to create internal tools tailored for their stack, using the provided void-builder and HOW_TO_CONTRIBUTE guides to customize UI and AI integrations.
Under The Hood
The Void Editor is an open-source code editor built on the VS Code platform, emphasizing extensibility, AI integration, and cross-platform portability. It combines TypeScript, Rust, and Electron to deliver a feature-rich yet customizable development environment.
Architecture
This project adopts a modular monorepo structure with well-defined layers for core functionality, extensions, and CLI tools. The architecture emphasizes service-oriented design and clear separation of concerns.
- Modular separation of concerns with distinct boundaries between editor components and extensions
- Layered approach separating UI, business logic, and system integration for maintainability
- Extensive use of Electron to enable cross-platform desktop deployment
- Strong support for portable and configurable installations across environments
Tech Stack
Built primarily with TypeScript and Node.js, the project leverages Rust for performance-critical operations and integrates modern web technologies.
- TypeScript as the primary language with ESM support and strong type safety
- Rust integration for CLI tools and performance-sensitive operations
- Electron framework enabling cross-platform desktop application capabilities
- Deep integration with VS Code extension ecosystem and configuration tools
Code Quality
Code quality is maintained through consistent patterns, comprehensive testing, and robust linting practices across the codebase.
- Extensive test coverage including unit and integration tests for core modules
- Strong error handling and graceful degradation in multi-environment setups
- Consistent naming conventions and modular structure for long-term maintainability
- Comprehensive documentation and contribution guidelines to support developer onboarding
What Makes It Unique
This project distinguishes itself through its innovative AI-focused features and flexible deployment strategies.
- Unique AI integration designed to enhance developer workflows in real-time
- Extensive support for portable installations and custom deployment configurations
- Automation of Linux distribution via AppImage creation scripts for seamless distribution
- Emphasis on extensibility and customization through modular architecture and plugin support