OSWorld 2.1
- Anthropic
- Reported in a later post or card
Results
| Model | Score | Source | Source check |
|---|---|---|---|
| Claude Sonnet 5.528 Sep 2026 · Anthropic | 80.1% | Introducing Claude Sonnet 5.5max effort; mean of 5 runs; September 10, 2026 task files | 1 other sourceThe same number is printed in 1 other lab document, linked below. |
| Claude Opus 5.522 Sep 2026 · Anthropic | 81.8% | Introducing Claude Opus 5.5max effort; mean of 5 runs; September 10, 2026 task files; the system card labels it OSWorld 2.0 | 3 other sourcesThe same number is printed in 3 other lab documents, linked below. |
| Claude Fable 5.11 Sep 2026 · Anthropic | 80.7% | Introducing Claude Opus 5.5max effort; mean of 5 runs; September 10, 2026 task files; the system card labels it OSWorld 2.0; reported later | 1 other sourceThe same number is printed in 1 other lab document, linked below. |
| Claude Opus 524 Jul 2026 · Anthropic | 74.0% | Introducing Claude Opus 5.5max effort; mean of 5 runs; September 10, 2026 task files; the system card labels it OSWorld 2.0; reported later | 1 other sourceThe same number is printed in 1 other lab document, linked below. |
| Claude Sonnet 530 Jun 2026 · Anthropic | 57.0% | Introducing Claude Sonnet 5.5max effort; mean of 5 runs; September 10, 2026 task files; reported later | 1 other sourceThe same number is printed in 1 other lab document, linked below. |
Benchmark
github.comHigher is better
XLANG LabIntroduced 16 Sep 2026