Benchmark
OSWorld 2.0
Compare computer-use agents on extended professional tasks.
Published leaderboard
Explore scores and evaluation settings on the publisher’s leaderboard.
Open leaderboardCoverage
108 tasks across seven professional domains and 21 subcategories, including research, creative production, engineering, finance, and healthcare.
Methodology
Agents operate desktop applications to complete tasks. The board distinguishes task sets, partial versus binary rewards, model settings, and resource use.