trycua/cua
Scale computer-use 2.0 with open-source drivers, cross-OS sandboxes, and benchmarks for training, evaluation, and data generation. observed · 2026-10-19
Health v2 · maintenance only
85/100
- Activity 100
- Release rhythm 88
- Longevity 44
How is this computed?
round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10) — computed 2026-10-21. Adoption (stars, forks) is never an input.
- gap_med: 0.0
- age_days: 627
- days_rel: 3
- days_push: 2
- n_releases_24m: 615
Adoption not part of the score
30149 stars · 2168 forks observed · 2026-10-19
What it is AI-extracted, prompt v1, taxonomy v1, 2026-10-05, confidence 0.93
Cua is an open‑source framework that provides drivers, sandbox environments, and fleet management for AI agents to interact with real desktops across macOS, Windows, Linux, and Android. It includes tools for automation, benchmarking, and data generation for computer‑use research.
Use cases
- run AI agents on real desktop applications for testing and data generation
- benchmark computer‑use agents across multiple operating systems
- scale parallel sandbox environments for large‑scale training workloads
- automate GUI interactions on macOS, Windows, or Linux without moving the cursor
- create reproducible computer‑use tasks and evaluate agent performance
When to choose
- you need a self‑hosted solution to run agents on actual OS desktops
- your project requires cross‑OS sandboxing and fleet scaling
- you want an open‑source benchmark suite for computer‑use agents
- you need a driver that can send clicks and keystrokes to background apps
When to avoid
- you only need simple headless script automation without GUI interaction
- your workload is limited to a single OS and does not require sandboxing
- you prefer a commercial cloud‑only solution with managed infrastructure
Facets
framework · maturity active
cli api-framework workflow-automation artificial-intelligence machine-learning computer-vision windows cloud python computer-use fleet sandbox driver cua-bench cua-s1 lume rust benchmark virtualization devops macos linux android docker
10 sources
- readme: https://github.com/trycua/cua · fetched 2026-10-19 · 096eb6be7cb3
- homepage: https://cua.ai · fetched 2026-08-29 · 51dd5251cc1a
- site_page: https://cua.ai/docs · fetched 2026-08-29 · b37a59dff341
- site_page: https://cua.ai/docs/how-to-guides · fetched 2026-08-29 · 542f1531a40e
- site_page: https://cua.ai/about · fetched 2026-08-29 · f8ef31f19346
- site_page: https://cua.ai/docs/how-to-guides/sandbox/scale-out · fetched 2026-08-29 · 9c1b4c11d49f
- site_page: https://cua.ai/docs/tutorials/your-first-cloud-fleet · fetched 2026-08-29 · e6ba9d251664
- site_page: https://cua.ai/docs/tutorials/drive-your-first-app · fetched 2026-08-29 · 08d1f713e66e
- site_page: https://cua.ai/docs/concepts/what-is-cua-bench · fetched 2026-08-29 · ea6564c2e077
- site_page: https://cua.ai/changelog · fetched 2026-08-29 · 7db073151523
Member repositories
| Repository | Role | Health v2 |
|---|---|---|
| trycua/cua | main | 85 |
For agents
Data as of 2026-10-19T08:05:51.212177+00:00 · Report a problem