HID-Chart - Biology: leaderboard
Metric: Accuracy (%; GPT-5 mini judge, 136 biology questions). Source: arxiv.org. Saturation forecast: Estimated already saturated. 16 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-4.1 | 61 |
| 2 | Gemini 2.0 Flash | 55.1 |
| 3 | GPT-4o | 45.6 |
| 4 | InternVL3-8B | 37.5 |
| 5 | Qwen 2.5 VL 7B | 37.5 |
| 6 | InternVL2.5-78B | 36 |
| 7 | Gemma 3 27B | 34.6 |
Interactive version: theaggregate.ai/benchmark?slug=hid-chart-biology · How It Works · Data refreshed daily, snapshot 2026-09-26.