OMIBench - Mathematics: leaderboard
OMIBench (2026) tests vision-language models on Olympiad-level science and mathematics problems whose evidence is spread across several images. This board is the GPTScore in percent on the mathematics problems; higher is better.
Metric: GPTScore (%, binarized LLM-judge equivalence of the final answer with the gold answer, micro-averaged) on the 430 mathematics problems of OMIBench (Olympiad-level multi-image problems from biology, chemistry, mathematics and physics competitions), zero-shot; higher is better. Source: arxiv.org. Saturation forecast: Around February 2027. 37 models tracked. Updated 7 Oct 2026.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3 Pro (Preview) | 62.56 | #119 |
| 2 | Qwen 3 VL 32B (Thinking) | 60.93 | #573 |
| 3 | Qwen 3 VL 30B A3B (Thinking) | 59.3 | #719 |
| 4 | GPT-5 Mini | 56.74 | #364 |
| 5 | O4 Mini | 56.61 | #368 |
| 6 | GPT-5 | 56.51 | #178 |
| 7 | Gemini 2.5 Pro | 53.49 | #276 |
| 8 | Qwen 3 VL 8B (Thinking) | 49.3 | #910 |
| 9 | Qwen 3 VL 235B A22B (Thinking) | 48.84 | #459 |
| 10 | Qwen 3 VL 4B (Thinking) | 44.65 | #1334 |
| 11 | Qwen 3 VL 32B Instruct | 40.7 | #606 |
| 12 | Gemini 2.5 Flash | 38.37 | #463 |
| 13 | Qwen 3 VL 235B A22B Instruct | 34.19 | #532 |
| 14 | Qwen 3 VL 30B A3B Instruct | 32.33 | #895 |
| 15 | InternVL3.5-8B | 32.33 | #1604 |
No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=omibench-mathematics · How It Works · Data refreshed daily, snapshot 2026-11-21.