OSWorld 2.1 offline - reported partial reward, step limit unspecified
Agents
Partial reward on the 82-task offline OSWorld 2.1 subset. Source step limit is unspecified; repeated-run averages do not license an item-level binomial noise model.
Top models (higher is better)
| Model | Score |
|---|---|
| Claude Haiku 5.5 | 72.4 |