What it is
Published Best in Computer Use (OSWorld) scores.
← Back to leaderboards
BENCHMARK RECORD
OSWorld-Verified
Source-linked result details from the TopLLM database.
BENCHMARK RECORD · 2026-09-15
OSWorld-Verified
What it tests
Published Best in Computer Use (OSWorld) scores.
How to read it
Scores are benchmark-specific and missing values are not estimated.
7 published entries85.0 top reported scorecomputer_use evaluation domain
Published leaders
Each bar is scaled only within this benchmark.
01Claude Fable 585.0Unresolved model record
02Claude Opus 4.883.4Unresolved model record
03Claude Sonnet 581.2Unresolved model record
04GPT-5.578.705Claude Sonnet 4.678.5Unresolved model record
06Kimi K2.673.107MiniMax M370.1Scores are benchmark-specific. Missing values remain N/A and are never inferred from another evaluation.
Compare in the benchmark matrix →