Back to Project
O

OpenGCUA

Project Oct 16, 2026

A research preview combining visual computer interaction, app APIs and code execution, with a preliminary macOS harness and Cross-Device SDK.

Visit primary link 2 sources attached
OpenGCUA authors' English stack diagram marking the agent design and trajectories released, macOS harness and SDK preliminary, and model, efficiency and evaluation work next.
Author-published research-preview diagram distinguishing released, preliminary and planned components. It is not an AITrending execution test. Rasterized and resized without cropping. Source: Tongyi-MAI / OpenGCUA authors
ai-agentsagent-ecosystem

What happened lately

OpenGCUA timeline

OpenGCUA is an agent harness and research stack, not a newly released foundation model. Its first preview opens an OSWorld integration example, evaluation trajectories from eight models, a preliminary real-device macOS harness and a Cross-Device SDK. The project’s own computer-and-browser model remains planned work.

The source harness requires macOS 14 or newer, Python 3.11-3.14, uv, Rust/Cargo 1.89 or newer and Xcode Command Line Tools; Python 3.12 is recommended. Users configure a model provider and API key separately and grant the driver Accessibility and Screen Recording access. The SDK documents adapters for macOS, Android, browsers and Linux. This does not establish equivalent native application availability on every platform.