# Cookie-Bench → Chess Puzzles

Calibration · https://benchgap.net/calibration/cookie-bench/current/to/chesspuzzles/current · data retrieved 2026-11-19

Chess Puzzles is estimated from Cookie-Bench with a linear curve fitted on 5 models measured on both: y = 3.0245·x + -2.1013, R² = 0.90, cross-validated error 8.0 pp. It is used for 2 estimates.

| Estimated model | Cookie-Bench | Chess Puzzles | Source |
|---|---|---|---|
| [Hy4](https://benchgap.net/model/hy4.md) | 88.8% | 58.4% | estimated ± 8.0 pp, low confidence |
| [ERNIE 5.5 Flash Preview](https://benchgap.net/model/ernie-5-5-flash-preview.md) | 88.2% | 56.6% | estimated ± 8.0 pp, low confidence |

## Fitted on

| Model | Cookie-Bench | Chess Puzzles |
|---|---|---|
| [Claude Opus 5.5](https://benchgap.net/model/claude-opus-5-5.md) | 91.5% | 67.0% |
| [DeepSeek V4.1 Flash 1017](https://benchgap.net/model/deepseek-v4-1-flash.md) | 87.9% | 46.0% |
| [GLM-5.3](https://benchgap.net/model/glm-5-3.md) | 80.4% | 37.0% |
| [GLM-5.3-Flash](https://benchgap.net/model/glm-5-3-flash.md) | 76.9% | 21.0% |
| [GPT-6.1 Sol](https://benchgap.net/model/gpt-6-1-sol.md) | 89.7% | 68.0% |
