# Terminal-Bench Hard → Terminal-Bench 2.1 (Vals AI)

Calibration · https://benchgap.net/calibration/1506 · data retrieved 2026-10-29

Terminal-Bench 2.1 (Vals AI) is estimated from Terminal-Bench Hard with a Hill curve fitted on 14 models measured on both: y = 0.5082 + (1.2000 − 0.5082)·x^6.00 / (0.66194^6.00 + x^6.00), R² = 0.80, cross-validated error 5.4 pp. It is used for 9 estimates.

| Estimated model | Terminal-Bench Hard | Terminal-Bench 2.1 (Vals AI) | Source |
|---|---|---|---|
| [GPT-5.4](https://benchgap.net/model/gpt-5-4.md) | 57.6% | 71.8% | estimated ± 5.4 pp, medium confidence |
| [Claude Opus 4.6](https://benchgap.net/model/claude-opus-4-6.md) | 48.5% | 60.1% | estimated ± 5.4 pp, medium confidence |
| [Claude Opus 4.5 Thinking](https://benchgap.net/model/claude-opus-4-5-thinking.md) | 47.0% | 58.7% | estimated ± 5.4 pp, medium confidence |
| [GPT-5.2](https://benchgap.net/model/gpt-5-2.md) | 47.0% | 58.7% | estimated ± 5.4 pp, medium confidence |
| [GPT-5.1](https://benchgap.net/model/gpt-5-1.md) | 45.5% | 57.4% | estimated ± 5.4 pp, medium confidence |
| [Muse Spark](https://benchgap.net/model/muse-spark.md) | 45.5% | 57.4% | estimated ± 5.4 pp, medium confidence |
| [Qwen 3.6 Max (preview)](https://benchgap.net/model/qwen3-6-max-preview.md) | 43.9% | 56.2% | estimated ± 5.4 pp, medium confidence |
| [GLM-5](https://benchgap.net/model/glm-5.md) | 43.2% | 55.8% | estimated ± 5.4 pp, medium confidence |
| [Solar Open 2](https://benchgap.net/model/solar-open-2.md) | 28.3% | 51.2% | estimated ± 5.4 pp, low confidence |

## Fitted on

| Model | Terminal-Bench Hard | Terminal-Bench 2.1 (Vals AI) |
|---|---|---|
| [Claude Fable 5](https://benchgap.net/model/claude-fable.md) | 62.9% | 80.5% |
| [Claude Opus 4.7](https://benchgap.net/model/claude-opus-4-7.md) | 54.5% | 68.5% |
| [Claude Opus 4.8](https://benchgap.net/model/claude-opus-4-8.md) | 58.3% | 71.9% |
| [Gemini 3.1 Pro](https://benchgap.net/model/gemini-3-1-pro.md) | 53.8% | 70.8% |
| [GLM-5.2](https://benchgap.net/model/glm-5-2.md) | 50.8% | 67.8% |
| [GPT-5.4 mini](https://benchgap.net/model/gpt-5-4-mini.md) | 52.3% | 54.7% |
| [GPT-5.5](https://benchgap.net/model/gpt-5-5.md) | 60.6% | 76.4% |
| [GPT-5.6 Sol](https://benchgap.net/model/gpt-5-6-sol.md) | 65.9% | 85.8% |
| [Kimi K2.6](https://benchgap.net/model/kimi-2-6.md) | 43.9% | 53.6% |
| [Kimi K2.7 Code](https://benchgap.net/model/kimi-k2-7-code.md) | 44.7% | 67.0% |
| [MiMo-V2.5-Pro](https://benchgap.net/model/mimo-v2-5-pro.md) | 43.2% | 57.3% |
| [Qwen3.6 Plus](https://benchgap.net/model/qwen3-6-plus.md) | 43.9% | 53.2% |
| [Qwen3.7 Max](https://benchgap.net/model/qwen3-7-max.md) | 50.8% | 61.0% |
| [Qwen3.7 Plus](https://benchgap.net/model/qwen3-7-plus.md) | 47.0% | 52.8% |