Void
Void
An open-source VS Code fork that connects directly to your chosen LLM without routing through a third-party backend.
Updated
The read
Void is a YC-backed open-source IDE forked from VS Code that lets developers wire their own LLM providers, whether cloud frontier models or self-hosted open-weight models, directly without going through Void's servers. It ships Tab autocomplete, inline editing, and a multi-mode chat covering Agent, Gather, and standard conversation. Because requests go from your machine straight to the provider, Void appeals to developers who want transparency over where their code travels. The project is still in beta and lacks the polish and community extensions of more mature tools.
Where it shines
- Direct LLM routing with no third-party backend
- Supports both frontier models and local open-weight models
- Fully open-source under MIT license
Worth knowing first
- Still in beta with incomplete feature parity to Cursor or Windsurf
- Small team means slower iteration on stability and UX
Index score
75.8
Provisional. This score is read from the product's public capability, not from the one-prompt rebuild the fully reviewed entries went through, so treat it as a placing rather than a verdict.
- Rank
- 26 of 54
- Builds
- Tools
- Category
- AI code editor
- Output
- Open-source VS Code fork with direct LLM connections
- Pricing
- Free tier
- Best for
- Privacy-conscious developers and self-hosters
Scores are our own editorial judgement, weighted craft 30, speed 25, control 25 and value 20.
Also worth a look
Full index- 02CursorCursorProvisionalThe editor most teams settled on, with an agent that ships whole features.AI code editorFree tier81.8
- 03ConductorConductorProvisionalA free macOS app for running multiple Claude Code or Codex agents in parallel isolated workspaces.AI code editorFree tier81.8
- 07Devin DesktopDevin DesktopProvisionalThe editor formerly called Windsurf, now a control room for several agents.AI code editorFree tier80.5
