← All benchmarks
OSWorld 2.1
Operating a real desktop through screenshots, mouse and keyboard to finish tasks across common apps.
How each maker reported it
- Claude Sonnet 5.5 Table 8.1.A, p.109
Claude results use adaptive thinking at max effort with default sampling, averaged over five trials. Competitor figures come from the developers' own published system cards or leaderboards.
Only each model's own maker's number is shown; scores a maker reports for rival models are left out.