# OpenGVLab/ScaleCUA

[ICLR 2026 Oral] ScaleCUA is the open-sourced computer use agents that can operate on cross-platform environments (Windows, macOS, Ubuntu, Android).

Repository: https://github.com/OpenGVLab/ScaleCUA
Canonical: https://ross.abutalabs.com/products/scalecua
Language: Python
License: Apache-2.0
License Family: permissive
Topics: data, gui-agents, models, computer-use-agents, online-evaluation-suite, scalecua
Last push: 2026-01-07T06:28:56+00:00

## Health v2 (maintenance only)
Score: 38/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 61, release rhythm 16, longevity 25
- inputs: {"age_days": 358, "days_push": 238, "days_rel": 349, "gap_med": null, "n_releases_24m": 1}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1133, forks 80 (observed 2026-08-28T04:03:42.955688+00:00)

## What it is
ScaleCUA is an open-source computer use agent project providing a large-scale cross-platform GUI operation dataset, trained models, and an online evaluation suite. It enables vision-language models to autonomously operate GUIs across Windows, macOS, Ubuntu, and Android.

## Use cases
- train a computer use agent to automate GUI tasks
- collect or use cross-platform GUI operation trajectory data
- evaluate GUI agents on OSWorld or WebArena benchmarks
- build an AI assistant that controls the desktop autonomously
- fine-tune a VLM for screen understanding and clicking
- benchmark cross-platform GUI automation models

## When to choose
- you need open training data or models for GUI automation agents
- you want a cross-platform computer use agent covering Windows, macOS, Linux, and Android
- you need a standardized evaluation suite for GUI agents

## When to avoid
- you only need simple scripted UI automation without learning-based agents
- you require a production-ready commercial RPA solution with support
- your target platform is not among the six supported operating systems

## Facets
- artifact type: dataset
- maturity: active
- function: machine-learning, agent-framework, computer-vision, data-generation, gui
- domain: artificial-intelligence, computer-vision, cross-platform
- platform: python, windows
- tags: computer-use-agents, gui-agents, vlm, screen-understanding, trajectory-data, benchmarking, iclr-2026, ai-agents, automation, macos, linux, android, gpu

## Member repositories
- OpenGVLab/ScaleCUA (main) score 38

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:03:42.955688+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T06:37:24.408440+00:00, confidence not recorded.
  - readme: https://github.com/OpenGVLab/ScaleCUA (fetched 2026-08-28T04:03:42.955688+00:00, sha e682ed6b360f)
- Data as of 2026-08-30T08:39:29.467469+00:00.