# UI-TARS-desktop

An open-source multimodal AI agent stack combining Agent TARS, a general-purpose GUI and vision agent driven from a CLI or web UI, with UI-TARS Desktop, a native desktop application built on the UI-TARS model.

Agent TARS is a general multimodal agent that brings GUI operation and vision into a terminal, computer or browser session, using multimodal LLMs together with Model Context Protocol tools; it ships primarily as a CLI and a web UI, and its CLI adds streaming output for tool calls, timing statistics, and an event-stream viewer for tracking data flow during a run.

UI-TARS Desktop is a separate desktop application providing a native GUI agent based on the UI-TARS model, with local and remote computer operators as well as a browser operator, so it can control a computer or browser directly without additional configuration.

It is written in TypeScript and licensed under Apache-2.0.

Website: https://github.com/bytedance/UI-TARS-desktop