# ResearchClawBench → OSWorld 2.0

Calibration · https://benchgap.net/calibration/1402 · data retrieved 2026-10-29

OSWorld 2.0 is estimated from ResearchClawBench with a offset logistic curve fitted on 5 models measured on both: y = 0.0719 + (1.2000 − 0.0719) / (1 + exp(−200.00·(x − 0.2211))), R² = 0.71, cross-validated error 5.1 pp. It is used for 3 estimates.

| Estimated model | ResearchClawBench | OSWorld 2.0 | Source |
|---|---|---|---|
| [GLM-5.2](https://benchgap.net/model/glm-5-2.md) | 20.7% | 13.6% | estimated ± 5.1 pp, low confidence |
| [MiMo-V2-Pro](https://benchgap.net/model/mimo-v2-pro.md) | 15.3% | 7.2% | estimated ± 5.1 pp, low confidence |
| [Grok 4.1](https://benchgap.net/model/grok-4-1.md) | 13.5% | 7.2% | estimated ± 5.1 pp, low confidence |

## Fitted on

| Model | ResearchClawBench | OSWorld 2.0 |
|---|---|---|
| [Claude Opus 4.7](https://benchgap.net/model/claude-opus-4-7.md) | 20.7% | 13.9% |
| [Claude Opus 4.8](https://benchgap.net/model/claude-opus-4-8.md) | 21.1% | 20.6% |
| [GPT-5.5](https://benchgap.net/model/gpt-5-5.md) | 17.0% | 13.0% |
| [Kimi K2.6](https://benchgap.net/model/kimi-2-6.md) | 18.0% | 4.6% |
| [MiniMax M3](https://benchgap.net/model/minimax-m3.md) | 19.8% | 4.6% |