Llama 3.2 11B Vision: benchmark results
Provider: Meta. Released 2024-09-25. Access: Open.
Aggregate Rating 1446 ± 21, rank #3937 of 10916 rated models, from 116 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| OODBench (In-Distribution) | 92.73 | Accuracy (%) on balanced yes/no existence questions ('Does t | 100 |
| OmniFood-Bench - Advisory (Kidney) | 46.11 | Advisory accuracy (%; share of images for which the model gi | 100 |
| OmniFood-Bench - Advisory (Obesity) | 36.18 | Advisory accuracy (%; share of images for which the model gi | 100 |
| TRAP - Text Privacy | 22 | Privacy score (%, 1 minus leakage) on the 100 text-document | 100 |
| Mind's Eye - Dynamic Structural Correspondence | 31.2 | Accuracy (%) on the dynamic structural correspondence task o | 88.2 |
| Mind's Eye - Hierarchical Pattern Equivalence | 29 | Accuracy (%) on the hierarchical pattern equivalence task of | 88.2 |
| Mind's Eye - Symmetric Structures | 29.3 | Accuracy (%) on the symmetric structures task of Mind's Eye | 88.2 |
| TRAP - Multimodal Privacy | 26.5 | Privacy score (%, 1 minus leakage) on the 200 multimodal-doc | 85.7 |
| TRAP - Privacy | 22.2 | Privacy score (%): 1 minus the leakage rate under the attack | 85.7 |
| Open Portuguese LLM - OAB Exams | 52.39 | Accuracy (%) | 81.7 |
| Open Portuguese LLM - ENEM | 71.59 | Accuracy (%) | 80.8 |
| OmniFood-Bench - Advisory (Diabetes) | 29.72 | Advisory accuracy (%; share of images for which the model gi | 80 |
Interactive version: theaggregate.ai/model?slug=llama-3-2-11b-vision · How It Works · Data refreshed daily, snapshot 2026-11-08.