trycua/cua
Scale computer-use 2.0 with open-source drivers, cross-OS sandboxes, and benchmarks for training, evaluation, and data generation.
307 best points · 230 comments · 13 Hacker News posts · 31,812 GitHub stars · View on GitHub
Hacker News discussions
- Apple Silicon and macOS VMs: Faster LLM Inference with llama.cpp (307 points, 41 comments, by frabonacci, 2026-08-11)
- Show HN: Drive any macOS app in the background without stealing the cursor (192 points, 43 comments, by frabonacci, 2026-04-28)
- Launch HN: Cua (YC X25) – Open-Source Docker Container for Computer-Use Agents (172 points, 73 comments, by frabonacci, 2025-04-23)
- Show HN: Lumier – Run macOS VMs in a Docker (159 points, 52 comments, by GreenGames, 2025-05-14)
- Show HN: CUA-S1 – A System One Model for Computer Use (95 points, 10 comments, by frabonacci, 2026-09-19)
- Show HN: Cua-Bench – a benchmark for AI agents in GUI environments (40 points, 8 comments, by someguy101010, 2026-01-26)
- When hackathon judging is a public benchmark: my report from Hack the North (18 points, 3 comments, by frabonacci, 2025-09-26)
- Show HN: Agent – A Local Computer-Use Operator for macOS (6 points, 0 comments, by f-trycua, 2025-03-30)
- macOS window internals: SkyLight enables multi-cursor background agents (4 points, 0 comments, by frabonacci, 2026-04-23)
- Windows computer-use: synthetic cursors for background agents (3 points, 0 comments, by frabonacci, 2026-05-27)
- Show HN: CuaBot – Co-op computer-use for any coding agent (2 points, 0 comments, by frabonacci, 2026-02-04)
- Cua's open-source alternative to Codex Computer History, encrypted locally (1 points, 0 comments, by frabonacci, 2026-08-18)
- A Compositor-Native Approach to Background Computer Use (1 points, 0 comments, by frabonacci, 2026-09-25)