# SimpleVQA → CountBench

Calibration · https://benchgap.net/calibration/1427 · data retrieved 2026-10-29

CountBench is estimated from SimpleVQA with a linear curve fitted on 8 models measured on both: y = 0.1457·x + 0.8526, R² = 0.34, cross-validated error 3.8 pp. It is used for 0 estimates.

## Fitted on

| Model | SimpleVQA | CountBench |
|---|---|---|
| [Claude Opus 4.5](https://benchgap.net/model/claude-opus-4-5.md) | 65.7% | 90.6% |
| [Gemini 3 Pro](https://benchgap.net/model/gemini-3-pro.md) | 73.2% | 97.3% |
| [GPT-5.2](https://benchgap.net/model/gpt-5-2.md) | 55.8% | 91.9% |
| [Kimi K2.5](https://benchgap.net/model/kimi-k2-5.md) | 71.2% | 94.1% |
| [LFM2.5-VL-3B](https://benchgap.net/model/lfm2-5-vl-3b.md) | 35.4% | 87.3% |
| [LFM3-VL-2B](https://benchgap.net/model/lfm3-vl-2b.md) | 36.3% | 93.0% |
| [Qwen3.5 397B](https://benchgap.net/model/qwen3-5-397b.md) | 67.1% | 97.2% |
| [Qwen3.6-27B](https://benchgap.net/model/qwen3-6-27b.md) | 56.1% | 97.8% |