# Llama 3.2 90B vs Step 3

> Step 3 is the stronger model overall, scoring 40.5 to 27.5 on the Noometry Index.

- Canonical page: https://noometry.com/compare/llama-3-2-90b-vs-step-3
- Last updated: 2026-10-12
- Shared benchmarks: 1

## Summary

- They share 1 benchmark with published results for both. Llama 3.2 90B scores higher in 0 categories and Step 3 in 4 categories; 4 gaps are clear of the uncertainty.
- The widest gap is in math, where Step 3 leads 37.6 to 11.1.

## Snapshot

| | Llama 3.2 90B | Step 3 |
|---|---|---|
| Provider | Meta | StepFun |
| Noometry Index | 27.5 | 40.5 |
| Rank | 331 | 149 |
| Context | — | — |
| Input $/M | — | — |
| Output $/M | — | — |
| Weights | Open | Open |

## Coding

- Llama 3.2 90B: —
- Step 3: 40.1 (#147)

| Benchmark | Llama 3.2 90B | Step 3 |
|---|---|---|
| LMArena Coding | — | 1367 |

## Agentic & Tool Use

- Llama 3.2 90B: 30.0 (#80)
- Step 3: —

| Benchmark | Llama 3.2 90B | Step 3 |
|---|---|---|
| BALROG | 27.3% | — |

## Reasoning

- Llama 3.2 90B: 21.7 (#217)
- Step 3: 28.4 (#105)

| Benchmark | Llama 3.2 90B | Step 3 |
|---|---|---|
| Kagi LLM Benchmark | — | 62.3% |
| EnigmaEval | 0.4% | — |
| LMArena Hard Prompts | — | 1355 |
| Epoch Capabilities Index | 125.5 | — |

## Math

- Llama 3.2 90B: 11.1 (#308)
- Step 3: 37.6 (#148)

| Benchmark | Llama 3.2 90B | Step 3 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 2.6% | — |
| LMArena Math | — | 1366 |
| MATH Level 5 | 39.4% | — |

## Knowledge

- Llama 3.2 90B: 21.7 (#274)
- Step 3: 36.8 (#164)

| Benchmark | Llama 3.2 90B | Step 3 |
|---|---|---|
| GPQA Diamond | 41% | — |
| LMArena Expert | — | 1333 |
| MMLU | 80.3% | — |

## Multimodal

- Llama 3.2 90B: 25.4 (#124)
- Step 3: 35.5 (#86)

| Benchmark | Llama 3.2 90B | Step 3 |
|---|---|---|
| LMArena Vision | 1000 | 1177 |
| GeoBench | 52% | — |

## Multilingual

- Llama 3.2 90B: —
- Step 3: 46.3 (#159)

| Benchmark | Llama 3.2 90B | Step 3 |
|---|---|---|
| LMArena Non-English | — | 1327 |
| LMArena Chinese | — | 1397 |
| LMArena German | — | 1371 |
| LMArena Korean | — | 1269 |
| LMArena Russian | — | 1331 |
| LMArena Spanish | — | 1371 |

## Instruction Following

- Llama 3.2 90B: —
- Step 3: 70.4 (#164)

| Benchmark | Llama 3.2 90B | Step 3 |
|---|---|---|
| LMArena Instruction Following | — | 1332 |

## Long Context

- Llama 3.2 90B: —
- Step 3: 40.3 (#157)

| Benchmark | Llama 3.2 90B | Step 3 |
|---|---|---|
| LMArena Longer Query | — | 1326 |

## Writing & Preference

- Llama 3.2 90B: —
- Step 3: 54.3 (#151)

| Benchmark | Llama 3.2 90B | Step 3 |
|---|---|---|
| LMArena Text | — | 1350 |
| LMArena Creative Writing | — | 1321 |
| LMArena Multi-Turn | — | 1341 |

## FAQ

### Is Llama 3.2 90B better than Step 3?

Step 3 is the stronger model overall, scoring 40.5 to 27.5 on the Noometry Index.

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

1 benchmark has published results for both models. Llama 3.2 90B has 9 scored results on Noometry and Step 3 has 17.