Skip to content
Bencher

TestsBenchmarks · Agentic

OSWorld 2.1 (strict pass rate)

Strict pass rate on the same tasks.

% solvedHigher is betterThe test's websiteOfficial page ↗

The best 3 · 3 models tested

Top 3 · best setting per model · 3 models, 3 results (0 independent)

Measured by independent testersIndependent Reported by the makerVendor-reported
OSWorld 2.1 (strict pass rate) leaderboard
Thinking levelSetting
1Claude Opus 5.5Anthropic48.7%Maxadaptive thinking, effort=maxMaker's own figureVendor-reportedAnthropic ↗—22 Sep 2026
2Claude Fable 5.1Anthropic42.8%Maxadaptive thinking, effort=maxMaker's own figureVendor-reportedAnthropic ↗—22 Sep 2026
3Claude Opus 5Anthropic37.2%Maxadaptive thinking, effort=maxMaker's own figureVendor-reportedAnthropic ↗—22 Sep 2026