OSWorld 2.0
Rank #32 of 51 · 5 models scored · top score 20.6
unknown
Strongest on contamination resistance, weakest on discrimination. Caveat: the public test set leaves contamination on the table.
Weighted composite. Each signal is computed from the source-backed leaderboard below.
Does it still separate models?
How far from ceiling / clustered at the top?
Public vs held-out; training-leak risk.
Apples-to-apples, or mixed / vendor-optimized?
How old is the benchmark?
| # | Model | Score | Evidence | Measured |
|---|---|---|---|---|
| 1 | Claude Opus 4.8 | 20.6 | unverified | 2026-05-28 |
| 2 | Claude Opus 4.7 | 18.2 | unverified | 2026-04-16 |
| 3 | GPT-5.5 | 13 | unverified | 2026-04-23 |
| 4 | Claude Sonnet 4.6 | 9.3 | unverified | 2026-02-17 |
| 5 | Kimi K2.6 | 4.6 | unverified | 2026-04-20 |