Self-hosted · Open source · Local LLMs
Your code. Your models.
Your workspace.
An IDE in your browser, powered by the models you choose. Edit code, ask an agent for help, review its changes, and work with Git and a terminal—all in one place.
Under active development toward 1.0. Try the current source build and share your feedback.
Bring your own models
Connect to Ollama, llama.cpp, or LM Studio. Use your own inference server for chat and coding tasks.
See what the agent changes
Let the agent read, edit, and run commands. Review diffs and choose the approval mode that fits how you work.
Pick up where you left off
Projects, conversations, and preferences stay with your deployment. Open it from another browser to continue working.
Make it your workspace
Work on folders on the host or choose a folder on your browser's device. Keep the same editor, chat, and review workflow across both.
Explore the architecture · Read the roadmap · Report a bug or ask a question