# HLE → Vals MMLU-Pro

Calibration · https://benchgap.net/calibration/hle/current/to/vals-mmlu-pro/current · data retrieved 2026-11-20

Vals MMLU-Pro is estimated from HLE with a linear curve fitted on 32 models measured on both: y = 0.1647·x + 0.7954, R² = 0.42, cross-validated error 2.4 pp. It is used for 6 estimates.

| Estimated model | HLE | Vals MMLU-Pro | Source |
|---|---|---|---|
| [Hy4](https://benchgap.net/model/hy4.md) | 59.1% | 89.3% | estimated ± 2.4 pp, medium confidence |
| [Apodex 1.1](https://benchgap.net/model/apodex-1-1.md) | 56.1% | 88.8% | estimated ± 2.4 pp, medium confidence |
| [Hy4 flash](https://benchgap.net/model/hy4-flash.md) | 55.2% | 88.6% | estimated ± 2.4 pp, medium confidence |
| [Agents-A1](https://benchgap.net/model/agents-a1.md) | 47.6% | 87.4% | estimated ± 2.4 pp, medium confidence |
| [Ling 3.1 Flash](https://benchgap.net/model/ling-3-1-flash.md) | 37.4% | 85.7% | estimated ± 2.4 pp, medium confidence |
| [MiniCPM5-2B](https://benchgap.net/model/minicpm5-2b.md) | 8.9% | 81.0% | estimated ± 2.4 pp, low confidence |

## Fitted on

| Model | HLE | Vals MMLU-Pro |
|---|---|---|
| [Claude Fable 5.1](https://benchgap.net/model/claude-fable-5-1.md) | 65.0% | 92.4% |
| [Claude Opus 4.5](https://benchgap.net/model/claude-opus-4-5.md) | 30.8% | 85.6% |
| [Claude Opus 4.7](https://benchgap.net/model/claude-opus-4-7.md) | 54.7% | 89.9% |
| [Claude Opus 4.8](https://benchgap.net/model/claude-opus-4-8.md) | 57.9% | 89.6% |
| [Claude Opus 5](https://benchgap.net/model/claude-opus-5.md) | 64.7% | 91.6% |
| [Claude Sonnet 4.6](https://benchgap.net/model/claude-sonnet-4-6.md) | 49.0% | 87.3% |
| [Claude Sonnet 5](https://benchgap.net/model/claude-sonnet-5.md) | 57.4% | 87.5% |
| [DeepSeek V4 Flash 0731](https://benchgap.net/model/deepseek-v4-flash-0731.md) | 34.8% | 86.2% |
| [DeepSeek V4 Pro](https://benchgap.net/model/deepseek-v4-pro.md) | 42.7% | 87.0% |
| [Gemini 3.5 Flash](https://benchgap.net/model/gemini-3-5-flash.md) | 40.2% | 89.5% |
| [GLM-4.7](https://benchgap.net/model/glm-4-7.md) | 24.8% | 82.7% |
| [GLM-5.1](https://benchgap.net/model/glm-5-1.md) | 31.0% | 86.9% |
| [GLM-5.2](https://benchgap.net/model/glm-5-2.md) | 40.5% | 86.7% |
| [GPT-5.4](https://benchgap.net/model/gpt-5-4.md) | 52.1% | 87.5% |
| [GPT-5.4 mini](https://benchgap.net/model/gpt-5-4-mini.md) | 41.5% | 84.6% |
| [GPT-5.4 nano](https://benchgap.net/model/gpt-5-4-nano.md) | 37.7% | 77.2% |
| [GPT-5.5](https://benchgap.net/model/gpt-5-5.md) | 52.2% | 88.1% |
| [Grok 4.3](https://benchgap.net/model/grok-4-3.md) | 35.0% | 85.8% |
| [Inkling](https://benchgap.net/model/inkling.md) | 46.0% | 86.3% |
| [Inkling-Small](https://benchgap.net/model/inkling-small.md) | 47.8% | 85.6% |
| [Kimi K2.6](https://benchgap.net/model/kimi-2-6.md) | 34.7% | 87.6% |
| [Kimi K3](https://benchgap.net/model/kimi-k3.md) | 56.0% | 88.0% |
| [Ling 3.0 Flash](https://benchgap.net/model/ling-3-0-flash.md) | 22.7% | 82.0% |
| [MiMo-V2.5-Pro](https://benchgap.net/model/mimo-v2-5-pro.md) | 48.0% | 84.6% |
| [MiniMax M2.7](https://benchgap.net/model/minimax-m2-7.md) | 28.0% | 80.4% |
| [Muse Spark](https://benchgap.net/model/muse-spark.md) | 50.4% | 87.3% |
| [Muse Spark 1.1](https://benchgap.net/model/muse-spark-1-1.md) | 62.1% | 88.7% |
| [Nemotron 3 Ultra](https://benchgap.net/model/nemotron-3-ultra.md) | 26.7% | 85.8% |
| [Qwen3.6 Plus](https://benchgap.net/model/qwen3-6-plus.md) | 28.8% | 87.7% |
| [Qwen3.7 Max](https://benchgap.net/model/qwen3-7-max.md) | 41.4% | 89.3% |
| [Qwen3.8-27B](https://benchgap.net/model/qwen3-8-27b.md) | 30.8% | 84.3% |
| [Qwen3.8 Max](https://benchgap.net/model/qwen3-8-max.md) | 43.6% | 88.6% |
