# Llama 3.2 90B vs Llama 3-70B

> Llama 3-70B is the stronger model overall, scoring 28.8 to 27.5 on the Noometry Index.

- Canonical page: https://noometry.com/compare/llama-3-2-90b-vs-llama-3-70b
- Last updated: 2026-10-12
- Shared benchmarks: 5

## Summary

- They share 5 benchmarks with published results for both. Llama 3.2 90B scores higher in 3 categories and Llama 3-70B in 1 category; 3 gaps are clear of the uncertainty.
- The widest gap is in agentic & tool use, where Llama 3.2 90B leads 30.0 to 21.1.
- The biggest single-benchmark swing is MATH Level 5: 39.4% for Llama 3.2 90B and 22.6% for Llama 3-70B.

## Snapshot

| | Llama 3.2 90B | Llama 3-70B |
|---|---|---|
| Provider | Meta | Meta |
| Noometry Index | 27.5 | 28.8 |
| Rank | 331 | 323 |
| Context | — | — |
| Input $/M | — | — |
| Output $/M | — | — |
| Weights | Open | Open |

## Coding

- Llama 3.2 90B: —
- Llama 3-70B: 35.8 (#218)

| Benchmark | Llama 3.2 90B | Llama 3-70B |
|---|---|---|
| BigCodeBench Instruct | — | 43.6% |
| LMArena Coding | — | 1206 |
| BigCodeBench Complete | — | 54.5% |
| HumanEval+ | — | 72% |
| MBPP+ | — | 69% |

## Agentic & Tool Use

- Llama 3.2 90B: 30.0 (#80)
- Llama 3-70B: 21.1 (#139)

| Benchmark | Llama 3.2 90B | Llama 3-70B |
|---|---|---|
| Cybench | — | 5% |
| BALROG | 27.3% | — |

## Reasoning

- Llama 3.2 90B: 21.7 (#217)
- Llama 3-70B: 18.0 (#288)

| Benchmark | Llama 3.2 90B | Llama 3-70B |
|---|---|---|
| Epoch Capabilities Index | 125.5 | 122.93 |
| Kagi LLM Benchmark | — | 35.1% |
| EnigmaEval | 0.4% | — |
| LMArena Hard Prompts | — | 1195 |
| DTBench | — | 54.2% |
| ForecastBench | — | 57.1 |
| WinoGrande | — | 83.5% |

## Math

- Llama 3.2 90B: 11.1 (#308)
- Llama 3-70B: 12.8 (#305)

| Benchmark | Llama 3.2 90B | Llama 3-70B |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 2.6% | 4.3% |
| MATH Level 5 | 39.4% | 22.6% |
| LMArena Math | — | 1218 |

## Knowledge

- Llama 3.2 90B: 21.7 (#274)
- Llama 3-70B: 20.8 (#277)

| Benchmark | Llama 3.2 90B | Llama 3-70B |
|---|---|---|
| GPQA Diamond | 41% | 40.6% |
| MMLU | 80.3% | 79.3% |
| LMArena Expert | — | 1149 |

## Multimodal

- Llama 3.2 90B: 25.4 (#124)
- Llama 3-70B: —

| Benchmark | Llama 3.2 90B | Llama 3-70B |
|---|---|---|
| LMArena Vision | 1000 | — |
| GeoBench | 52% | — |

## Multilingual

- Llama 3.2 90B: —
- Llama 3-70B: 33.6 (#251)

| Benchmark | Llama 3.2 90B | Llama 3-70B |
|---|---|---|
| LMArena Non-English | — | 1142 |
| LMArena Chinese | — | 1114 |
| LMArena French | — | 1232 |
| LMArena German | — | 1169 |
| LMArena Japanese | — | 1017 |
| LMArena Korean | — | 1017 |
| LMArena Russian | — | 1159 |
| LMArena Spanish | — | 1241 |

## Instruction Following

- Llama 3.2 90B: —
- Llama 3-70B: 62.5 (#238)

| Benchmark | Llama 3.2 90B | Llama 3-70B |
|---|---|---|
| LMArena Instruction Following | — | 1194 |

## Long Context

- Llama 3.2 90B: —
- Llama 3-70B: 35.6 (#240)

| Benchmark | Llama 3.2 90B | Llama 3-70B |
|---|---|---|
| LMArena Longer Query | — | 1174 |

## Writing & Preference

- Llama 3.2 90B: —
- Llama 3-70B: 42.8 (#231)

| Benchmark | Llama 3.2 90B | Llama 3-70B |
|---|---|---|
| LMArena Text | — | 1221 |
| LMArena Creative Writing | — | 1210 |
| LMArena Multi-Turn | — | 1223 |

## FAQ

### Is Llama 3.2 90B better than Llama 3-70B?

Llama 3-70B is the stronger model overall, scoring 28.8 to 27.5 on the Noometry Index.

### How many benchmarks do Llama 3.2 90B and Llama 3-70B share?

5 benchmarks have published results for both models. Llama 3.2 90B has 9 scored results on Noometry and Llama 3-70B has 31.