Autonomous AI coding that runs for hours while you do other things.
A command-line interface for multi-pass development workflows. Use any model, any provider, on your terms. No telemetry, no lock-in, fully open source.
npm install -g @vybestack/llxprt-code
Works with Claude, GPT, Gemini, and local models. Switch providers mid-session, use multi-account failover, or load balance across API keys.
Not just one-shot answers. LLxprt plans, implements, tests, debugs, and iterates—across hours, not minutes.
Use your existing Claude, OpenAI, or Google subscriptions via OAuth. No separate billing, no middleman markup.
Your code stays on your machine. No usage data sent to us. Privacy by design.
Apache 2.0 licensed. Community-driven development. Fork it, extend it, make it yours.
Built for developers who live in the command line. Vim keybindings, shell integration, pipe-friendly output.
Get a coding partner that actually understands your codebase. Run long refactoring tasks while you focus on architecture.
Standardize on one tool, use multiple providers. When one API is down or throttled, fail over automatically.
No telemetry means no data leaves your environment. Perfect for regulated industries and security-focused teams.
Active development with 340+ community members and growing.
