GeoSym-Bench: leaderboard
Large Multimodal Models (LMMs) often struggle with geometric reasoning due to visual hallucinations and a lack of mathematically precise Chain-of-Thought (CoT) data. To address this, we propose the GeoSym Engine, an automated and scalable neuro-symbolic framework. By leveraging a type-conditional grammar and an analytic SymGT Solver, it derives exact symbolic ground truths and seamlessly integrate
Metric: Overall (self-reported). Source: benchmarklist.com. Saturation forecast: Around May 2027. 22 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Qwen 3.5 397B A17B | 87.47 | #146 |
| 2 | Gemini 3 Pro | 79.51 | #80 |
| 3 | Qwen 3 VL 235B A22B | 77.34 | #238 |
| 4 | GPT-5 | 76.55 | #98 |
| 5 | Seed 1.8 | 73.42 | #139 |
| 6 | Qwen 3 VL 8B Instruct | 55.94 | #420 |
| 7 | Qwen 3 VL 4B Instruct | 52.39 | #515 |
| 8 | Qwen 2.5 VL 7B Instruct | 39.19 | #654 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=geosym-bench · How It Works · Data refreshed daily, snapshot 2026-10-20.