Llama 3.2 90B
Best configuration ranks #561 of 610 on the BenchLeader Index at 37.5 ±5.9. Last measured 7 Mar 2025stale: no new result in six months. Released 24 Sept 2024.
- Blended price
- –
- Output speed
- –
- First answer
- –
- Context
- 128k
- Overall index38
- Reasoning32
- Maths33
- Multimodal23
- Composite37
Benchmark results
One column per reasoning effort. Rank is among every configuration of every model on that benchmark. Hover a score for the run it came from.
Reasoning
| Benchmark | default | Source |
|---|---|---|
| GPQA Diamond | 41.0%#227 | Epoch AI Benchmarking Hub |
| GPQA Diamond (AA)not in index | 43.2%#437 | Artificial Analysis |
| Humanity's Last Exam (AA)not in index | 4.5%#419 | Artificial Analysis |
Maths
| Benchmark | default | Source |
|---|---|---|
| OTIS Mock AIME | 2.6%#235 | Epoch AI Benchmarking Hub |
| MATH Level 5 | 39.4%#61 | Epoch AI Benchmarking Hub |
Multimodal
| Benchmark | default | Source |
|---|---|---|
| MMMU-Pro | 39.5%#231 | Artificial Analysis |
| VISTA | 24.6%#55 | Scale AI SEAL |
Composite
| Benchmark | default | Source |
|---|---|---|
| AA Intelligence Index | 6.4#446 | Artificial Analysis |
See also
Data as of 9 Sept 2026. Compare with another model.