Aya Vision 32B: benchmark results
Provider: Cohere. Released 2025-03-04. Access: Open.
Aggregate Rating 1501 ± 31, rank #2349 of 10916 rated models, from 42 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Cedar - Multimodal - Chinese | 43.14 | Accuracy (%) | 100 |
| MUNIChus - French | 18.12 | CIDEr of the zero-shot French news image caption generated f | 100 |
| MUNIChus - Indonesian | 24.94 | CIDEr of the zero-shot Indonesian news image caption generat | 100 |
| KOFFVQA - Recognition | 100 | Score (%) | 98.8 |
| Cedar - Text-Only - Chinese | 43.32 | Accuracy (%) | 86.7 |
| KOFFVQA - Relationship | 70 | Score (%) | 77.8 |
| Cedar - Text-Only - Hindi | 46.26 | Accuracy (%) | 75 |
| Cedar - Text-Only - Japanese | 36.57 | Accuracy (%) | 73.3 |
| Cedar - Multimodal - Arabic | 38.74 | Accuracy (%) | 69.2 |
| Cedar - Multimodal - Japanese | 31.01 | Accuracy (%) | 69.2 |
| MUNIChus - Chinese | 25.13 | CIDEr of the zero-shot Chinese news image caption generated | 66.7 |
| MUNIChus - English | 35.55 | CIDEr of the zero-shot English news image caption generated | 66.7 |
Interactive version: theaggregate.ai/model?slug=aya-vision-32b · How It Works · Data refreshed daily, snapshot 2026-11-08.