AI benchmarks and leaderboards are different things
Learn the difference between a benchmark, an evaluation run and a leaderboard, and use each to answer the right question.
Model reports
OSWorld
Screen resolution, OS image and action space all change results. The Verified and 2.0 revisions are not comparable to the original.
Evaluation context and evidence
- Source
- Model reports
- Publisher
- Not recorded
- Benchmark release
- 2024-04-11
- Collected
- Not recorded
- Modality
- Not recorded
These are source-reported observations, not independent reruns by this site. Check the task version, evaluation protocol, unit and reporting date before comparing models. Results from different setups are not a universal ranking.
Original sources
Reported results 4
Shown in source observation order, not ranked. Dates retain the precision provided by the source.
| Model | Value | Unit | Protocol | Reported date and basis | Source |
|---|---|---|---|---|---|
| Claude Opus 5 | 70.57 | percent | max effort, first-attempt success rate, averaged over 5 runs, 1080p, max 500 action steps | 2026-07-24 · document_publication | Claude Opus 5 |
| Kimi K3 | 58.3 | percent | reasoning=max | 2026-06-13 · document_publication | Kimi K3 |
| Kimi K3 | 84.8 | percent | reasoning=max | 2026-06-13 · document_publication | Kimi K3 |
| Qwen3.5-397B-A17B | 62.2 | percent | thinking (temp 0.6, top_p 0.95, top_k 20) | 2026-02-16 · document_publication | Qwen3.5-397B-A17B |