PUMA Polish Multimodal - Vision Mean: leaderboard
Unweighted mean of PUMA's six vision categories (History and culture, Contemporary life, Geography and environment, OCR, Document QA, Structured extraction), the Mean score of the paper's Table 5 (arXiv 2608.21853v1). Strict score 0-100, higher is better, from the OPI-PIB/puma Space data.json; only models scored on all six are listed.
Metric: Mean Strict Accuracy over 6 Vision Categories (%). Source: huggingface.co. Saturation forecast: Around February 2027. 80 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 4 Argon | 84.83 | #9 |
| 2 | Gemini 3.8 Flash | 84 | #30 |
| 3 | Gemini 3.7 Flash | 84 | #24 |
| 4 | Gemini 3.1 Pro (Preview) | 82.17 | #64 |
| 5 | Gemini 3.6 Flash | 81.33 | #57 |
| 6 | Gemini 3.5 Flash | 81.17 | #66 |
| 7 | Muse Wonder | 80.5 | #18 |
| 8 | Claude Fable 5 | 78 | #20 |
| 9 | GPT-6 Astra | 77.33 | #6 |
| 10 | GPT-6.1 Sol | 76.33 | #7 |
| 11 | Claude Opus 5.5 | 75.33 | #3 |
| 12 | Muse Spark 1.2 | 73.33 | #69 |
| 13 | GPT-5.5 | 73.17 | #33 |
| 14 | Gemini 2.5 Pro | 72.33 | #159 |
| 15 | GPT-5.6 Sol | 71.67 | #22 |
No result here: #10 Claude Fable 5.1.
Interactive version: theaggregate.ai/benchmark?slug=puma-polish-multimodal-vision-mean · How It Works · Data refreshed daily, snapshot 2026-11-05.