# OmniDocBench 1.5 → MathVision

Calibration · https://benchgap.net/calibration/omnidocbench15/current/to/mathvision/current · data retrieved 2026-11-21

MathVision is estimated from OmniDocBench 1.5 with a Hill curve fitted on 8 models measured on both: y = 0.0000 + (1.2000 − 0.0000)·x^6.00 / (0.76375^6.00 + x^6.00), R² = 0.55, cross-validated error 4.5 pp. It is used for 0 estimates.

## Fitted on

| Model | OmniDocBench 1.5 | MathVision |
|---|---|---|
| [Claude Opus 4.5](https://benchgap.net/model/claude-opus-4-5.md) | 87.7% | 74.3% |
| [Gemini 3 Pro](https://benchgap.net/model/gemini-3-pro.md) | 88.5% | 86.6% |
| [GPT-5.2](https://benchgap.net/model/gpt-5-2.md) | 85.7% | 83.0% |
| [Kimi K2.5](https://benchgap.net/model/kimi-k2-5.md) | 88.8% | 84.2% |
| [Qwen3.5 397B](https://benchgap.net/model/qwen3-5-397b.md) | 90.8% | 88.6% |
| [Qwen3.7 Plus](https://benchgap.net/model/qwen3-7-plus.md) | 91.4% | 90.3% |
| [Qwen3.8-27B](https://benchgap.net/model/qwen3-8-27b.md) | 91.1% | 90.0% |
| [Qwen3.8 Max](https://benchgap.net/model/qwen3-8-max.md) | 92.1% | 95.2% |
