New: Agent Index for knowledge work is live.Explore the index ↗
Compare models

Best LLMs for computer use

OSWorld

OSWorld evaluates computer-use agents on practical tasks in a desktop environment. It tests whether agents can interact with applications and complete multi-step work through a graphical interface. Results below are restricted to the stored 2.1 full-set evaluation.

Updated: Data as of:
Limited reportingNot saturated
Official benchmark
OSWorld results. Each row identifies its benchmark version and evaluation track. Scores from different versions are not directly comparable. Best effort from the latest stored report date per model and version.
Model / creatorVersionReported score
Claude Opus 5Anthropic · Proprietary2.1 · Full set44.33%
Muse Spark 1.3Meta · Proprietary2.1 · Full set39.25%
Muse Spark 1.2Meta · Proprietary2.1 · Full set15.84%
View all 3 published reports