Open LMM Reasoning - DynaMath - Subject-plane geometry: leaderboard
The "DynaMath - Subject-plane geometry" subtask of Open LMM Reasoning. A task from the Open LMM Reasoning leaderboard, which evaluates large multimodal models on multi-step visual, mathematical and logical reasoning problems.
Metric: Accuracy (%). Source: huggingface.co. Saturation forecast: Estimated already saturated. 100 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Doubao-1.5-Pro | 46.8 | #287 |
| 2 | GPT-4.1 Mini | 44.2 | #355 |
| 3 | GPT-4o ChatGPT | 44.2 | |
| 4 | Gemini 2.0 Flash | 40.3 | #342 |
| 5 | InternVL3-78B | 39 | #352 |
| 6 | InternVL2.5-78B-CoT | 39 | #469 |
| 7 | Claude 3.7 Sonnet | 37.7 | #249 |
| 8 | Gemma 3 27B | 37.7 | #607 |
| 9 | InternVL3-38B | 36.4 | #405 |
| 10 | GPT-4.1 (2025-04-14) | 36.4 | #250 |
| 11 | InternVL3-8B | 33.8 | #620 |
| 12 | Gemini 1.5 Pro (002) | 33.8 | #349 |
| 13 | InternVL3-14B | 33.8 | #505 |
| 14 | Claude 3.5 Sonnet (20241022) | 31.2 | #291 |
| 15 | Gemma 3 12B | 31.2 | #671 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=open-lmm-reasoning-dynamath-subject-plane-geometry · How It Works · Data refreshed daily, snapshot 2026-10-20.