CUA-Gym is an open-source pipeline and dataset for generating and verifying training environments and tasks for computer-use reinforcement learning agents. It provides thousands of RLVR tuples and supports research in agent training and evaluation across desktop and web applications.
CUA-Gym is an AI & ML product. It focuses on scaling and verifying training environments for reinforcement learning agents that interact with computer applications. CUA-Gym is an open-source project aimed at AI researchers and RL developers. The project is open source (Open Source). CUA-Gym is available on the web, and it can be self-hosted.
XLang Lab builds and maintains CUA-Gym, and the product first shipped in 2026. The project is developed in the open on GitHub with 167 stars and 31 commits in the last 90 days. Across PulseGate's embedding index, CUA-Gym has few near neighbours, marking it as relatively distinct. Among its 5 catalogued features are RL environment generation, verifiable tasks, and open dataset.
Latest indexed changes and source events
xlang-ai/CUA-Gym discovered by the PulseGate indexer
Other apps tracked under the same category.
