# ResearchClawBench → OSWorld-Verified

Calibration · https://benchgap.net/calibration/1401 · data retrieved 2026-10-29

OSWorld-Verified is estimated from ResearchClawBench with a Hill curve fitted on 9 models measured on both: y = 0.0000 + (0.7934 − 0.0000)·x^6.00 / (0.10963^6.00 + x^6.00), R² = 0.55, cross-validated error 5.5 pp. It is used for 2 estimates.

| Estimated model | ResearchClawBench | OSWorld-Verified | Source |
|---|---|---|---|
| [Grok 4.1](https://benchgap.net/model/grok-4-1.md) | 13.5% | 61.7% | estimated ± 5.5 pp, low confidence |
| [Grok 4.3](https://benchgap.net/model/grok-4-3.md) | 12.4% | 53.7% | estimated ± 5.5 pp, low confidence |

## Fitted on

| Model | ResearchClawBench | OSWorld-Verified |
|---|---|---|
| [Claude Opus 4.6](https://benchgap.net/model/claude-opus-4-6.md) | 19.9% | 72.7% |
| [Claude Opus 4.8](https://benchgap.net/model/claude-opus-4-8.md) | 21.1% | 83.4% |
| [Gemini 3.5 Flash](https://benchgap.net/model/gemini-3-5-flash.md) | 18.0% | 78.4% |
| [GPT-5.4](https://benchgap.net/model/gpt-5-4.md) | 15.3% | 75.0% |
| [GPT-5.5](https://benchgap.net/model/gpt-5-5.md) | 17.0% | 78.7% |
| [Kimi K2.6](https://benchgap.net/model/kimi-2-6.md) | 18.0% | 73.1% |
| [Kimi K2.5](https://benchgap.net/model/kimi-k2-5.md) | 14.0% | 63.3% |
| [MiniMax M3](https://benchgap.net/model/minimax-m3.md) | 19.8% | 70.1% |
| [Qwen3.5 397B](https://benchgap.net/model/qwen3-5-397b.md) | 14.2% | 62.2% |