PUMA Polish Multimodal - Geography and Environment: leaderboard
Geography and environment category of PUMA (Polish Unified Multimodal Assessment), a culturally grounded Polish benchmark for vision-language models (Images group; arXiv 2608.21853v1). Strict score, the percentage of answers meeting every verification rule, 0-100, higher is better, from the OPI-PIB/puma Space data.json; audio-pipeline rows and the auxiliary Soft score are excluded.
Metric: Strict Accuracy (%). Source: huggingface.co. Saturation forecast: Around February 2027. 80 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3.8 Flash | 91 | #30 |
| 2 | Gemini 3.7 Flash | 89 | #24 |
| 3 | Gemini 4 Argon | 89 | #9 |
| 4 | Gemini 3.5 Flash | 86 | #66 |
| 5 | Gemini 3.1 Pro (Preview) | 85 | #64 |
| 6 | GPT-6 Astra | 85 | #6 |
| 7 | Gemini 3.6 Flash | 85 | #57 |
| 8 | Claude Fable 5 | 85 | #20 |
| 9 | Claude Opus 5.5 | 82 | #3 |
| 10 | Muse Wonder | 82 | #18 |
| 11 | GPT-6.1 Sol | 81 | #7 |
| 12 | GPT-5.6 Sol | 78 | #22 |
| 13 | GPT-5.5 | 77 | #33 |
| 14 | Gemini 2.5 Pro | 72 | #159 |
| 15 | GPT-5.6 Terra | 72 | #40 |
No result here: #10 Claude Fable 5.1.
Interactive version: theaggregate.ai/benchmark?slug=puma-polish-multimodal-geography-and-environment · How It Works · Data refreshed daily, snapshot 2026-11-05.