OSWorld-Verified (BenchmarkList Multi-Source): leaderboard
BenchmarkList combines reported desktop computer-use success rates from official OSWorld submissions and vendor evaluations. Agents operate real desktop applications and are graded on the resulting task state. Task populations, step budgets and agent harnesses are not uniformly bound across the imported rows, so this cohort remains separate from the primary OSWorld-Verified board.
Metric: Success rate (imported). Source: benchmarklist.com. Saturation forecast: Around December 2027. 143 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Qwen 3.8 Max | 86.1 | #33 |
| 2 | Claude Fable 5 | 86 | #17 |
| 3 | Kimi K3 | 84.8 | #32 |
| 4 | Qwen 3.8 27B | 84.3 | #111 |
| 5 | Claude Opus 4.8 | 83.4 | #36 |
| 6 | Claude Opus 5 | 83.4 | #13 |
| 7 | GPT-5.6 Sol | 83 | #20 |
| 8 | Gemini 3.6 Flash | 83 | #52 |
| 9 | Claude Opus 4.7 | 82.8 | #51 |
| 10 | MiMo-V2.6-Pro | 82 | #42 |
| 11 | Claude Sonnet 5 | 81.2 | #60 |
| 12 | MiMo-V2.6-Flash | 80.8 | #72 |
| 13 | Muse Spark 1.1 | 80.8 | #55 |
| 14 | GPT-5.5 | 78.7 | #30 |
| 15 | Gemini 3.5 Flash | 78.4 | #62 |
No result here: #4 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=osworld-verified-benchmarklist-multi-source · How It Works · Data refreshed daily, snapshot 2026-10-24.