Local Studio

Run your intelligence
at home.

AI workspace built to help you go local.

Download for macOSSetup
Local Studio workbench showing a coding agent and a local document in the integrated browser.

All in one

Local AI is hard, this will make it easier.

Local Studio keeps the model, the runtime, and the hardware together. Nothing important disappears behind a provider abstraction.

01

Control

Local and remote controllers, live status, launch state, logs, and metrics.

Local Studio usage screen showing proxied tokens, requests, sessions, active days, and token activity.
02

Serve

vLLM, SGLang, MLX, and llama.cpp behind one OpenAI-compatible surface.

Local Studio model configuration screen showing searchable Hugging Face models, hardware fit, and downloads.
03

Work

Models, providers, browser, files, terminal, and agents in the same session.

Local Studio workbench with agent reasoning, terminal output, repository changes, and tool activity.

KittyLitter

Your agents, in your pocket.

Pair your phone from Local Studio, then keep working with Codex, Claude, OpenCode, Pi, and Droid from anywhere.

See how mobile worksGet KittyLitter
Light
Local Studio phone connection screen in light mode with a KittyLitter iPhone capture covering the private pairing code.
Dark
Local Studio phone connection screen in dark mode with a KittyLitter iPhone capture covering the private pairing code.

Local Studio

Your models. Your hardware. One surface.

Download for macOSView on GitHubSetup