Kaggle FACTS Multimodal: leaderboard
FACTS Multimodal sub-task evaluating factual accuracy when reasoning over both text and image inputs from source documents.
Metric: Score (%). Source: www.kaggle.com. Saturation forecast: Around 2029. 38 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3.8 Flash | 50.7 | #21 |
| 2 | Gemini 3.7 Flash | 49.83 | #23 |
| 3 | GPT-6.1 Sol | 48.46 | #7 |
| 4 | Claude Opus 5.5 | 48.37 | #3 |
| 5 | GPT-6 Astra | 47.73 | #6 |
| 6 | Gemini 3.6 Flash | 47.21 | #51 |
| 7 | Gemini 2.5 Pro | 46.86 | #152 |
| 8 | GPT-5.5 | 45.77 | #31 |
| 9 | GPT-5 | 44.08 | #98 |
| 10 | GPT-6 Sol | 43.06 | #14 |
| 11 | GPT-5.1 | 41.8 | #135 |
| 12 | GPT-5 Mini | 41.47 | #186 |
| 13 | Gemini 3.5 Flash | 41.44 | #60 |
| 14 | Gemini 3.1 Pro (Preview) | 41.34 | #58 |
| 15 | Gemini 3 Flash (Preview) | 41.28 | #82 |
No result here: #8 Gemini 4 Argon, #9 Claude Fable 5.1, #10 Muse Wonder Full.
Interactive version: theaggregate.ai/benchmark?slug=kaggle-facts-multimodal · How It Works · Data refreshed daily, snapshot 2026-10-20.