Loading...
Loading...
Available on 1 platform
Sign in to view source links and access this dataset
OSWorld-Verified Model Trajectories contains between 100,000 and 1,000,000 evaluation records of multimodal AI agents performing tasks in real computer environments. Created by xlangai and updated in March 2026, the data captures verified execution paths and screenshots from state-of-the-art models tested on the OSWorld benchmark.
The dataset is distributed as a series of zip files containing screenshots and trajectory logs; users should ensure they have sufficient disk space for extraction.