Frontier Log

Benchmark · Active

OSWorld 2.1

0102030405060Jul 2026Oct 202625.6%Sonnet 548.7%Opus 5.543.5%Sonnet 5.50102030405060Jul 2026Oct 202625.6%Sonnet 548.7%Opus 5.543.5%Sonnet 5.5

Results

ModelScoreSourceSource check
Claude Sonnet 5.528 Sep 2026 · Anthropic43.5%Claude Sonnet 5.5 System Cardmax effort; mean of 5 runs; September 10, 2026 task files
Source checkedA person checked this number against its source on 20 Oct 2026. This confirms the number matches what the lab published, not that the score was reproduced.
Claude Opus 5.522 Sep 2026 · Anthropic48.7%Claude Sonnet 5.5 System Cardmax effort; mean of 5 runs; September 10, 2026 task files; reported later
Source checkedA person checked this number against its source on 20 Oct 2026. This confirms the number matches what the lab published, not that the score was reproduced.
Claude Sonnet 530 Jun 2026 · Anthropic25.6%Claude Sonnet 5.5 System Cardmax effort; mean of 5 runs; September 10, 2026 task files; reported later
Source checkedA person checked this number against its source on 20 Oct 2026. This confirms the number matches what the lab published, not that the score was reproduced.

Benchmark

github.comHigher is better

XLANG LabIntroduced 16 Sep 2026

All benchmarks