Llama 3.2 90B: benchmark results
Meta's 90B vision-language base model from Llama 3.2, adding cross-attention image adapters to Llama text weights with a 128K context. Provider: Meta. Released 2024-09-25. Access: Open.
Unified ELO 1533 ± 1, rank #526 of 1392 rated models, from 22 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Translation Set1 to en spBleu | 43.7 | Set1→en spBLEU (self-reported) | 75 |
| AI for Education Pedagogy - Technology | 82.08 | Accuracy (%) | 69.1 |
| MMMU Benchmark | 60.3 | Validation Score | 66.1 |
| AI for Education Pedagogy - Maths | 76.19 | Accuracy (%) | 47.9 |
| AI for Education Pedagogy | 76.31 | Accuracy (%) | 45.4 |
| AI for Education Pedagogy - Secondary | 75 | Accuracy (%) | 44.3 |
| AI for Education Pedagogy - Social studies | 74.55 | Accuracy (%) | 43.9 |
| AI for Education SEND | 70.64 | Accuracy (%) | 41.3 |
| AI for Education Pedagogy - Primary | 79.81 | Accuracy (%) | 39.7 |
| AI for Education Pedagogy - Science | 74.86 | Accuracy (%) | 35.7 |
| Translation Set1 to en COMET22 | 88.5 | Set1→en COMET22 (self-reported) | 29.2 |
| AI for Education Visual Maths - Geometry | 36.92 | Accuracy (%) | 27.7 |
Interactive version: theaggregate.ai/model?slug=llama-3-2-90b · How It Works · Data refreshed daily, snapshot 2026-09-05.