# Llama 3-70B vs Qwen3 32B

> Qwen3 32B is the stronger model overall, scoring 39.2 to 28.8 on the Noometry Index.

- Canonical page: https://noometry.com/compare/llama-3-70b-vs-qwen3-32b
- Last updated: 2026-10-12
- Shared benchmarks: 18

## Summary

- They share 18 benchmarks with published results for both. Llama 3-70B scores higher in 0 categories and Qwen3 32B in 9 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in math, where Qwen3 32B leads 39.7 to 12.8.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 4.3% for Llama 3-70B and 66.9% for Qwen3 32B.

## Snapshot

| | Llama 3-70B | Qwen3 32B |
|---|---|---|
| Provider | Meta | Alibaba (Qwen) |
| Noometry Index | 28.8 | 39.2 |
| Rank | 323 | 172 |
| Context | — | 131K |
| Input $/M | — | $0.70 |
| Output $/M | — | $2.80 |
| Weights | Open | Open |

## Coding

- Llama 3-70B: 35.8 (#218)
- Qwen3 32B: 37.7 (#190)

| Benchmark | Llama 3-70B | Qwen3 32B |
|---|---|---|
| LMArena Coding | 1206 | 1358 |
| Aider Polyglot | — | 40% |
| SciCode | — | 35.4% |
| BigCodeBench Instruct | 43.6% | — |
| BigCodeBench Complete | 54.5% | — |
| HumanEval+ | 72% | — |
| MBPP+ | 69% | — |

## Agentic & Tool Use

- Llama 3-70B: 21.1 (#139)
- Qwen3 32B: 32.6 (#62)

| Benchmark | Llama 3-70B | Qwen3 32B |
|---|---|---|
| Berkeley Function Calling Leaderboard | — | 48.7% |
| Cybench | 5% | — |

## Reasoning

- Llama 3-70B: 18.0 (#288)
- Qwen3 32B: 20.2 (#241)

| Benchmark | Llama 3-70B | Qwen3 32B |
|---|---|---|
| Kagi LLM Benchmark | 35.1% | 54.9% |
| LMArena Hard Prompts | 1195 | 1334 |
| DTBench | 54.2% | 67.5% |
| Epoch Capabilities Index | 122.93 | 138.51 |
| CritPt | — | 0.3% |
| Chess Puzzles | — | 5% |
| LMCA | — | 17.3% |
| ForecastBench | 57.1 | — |
| WinoGrande | 83.5% | — |

## Math

- Llama 3-70B: 12.8 (#305)
- Qwen3 32B: 39.7 (#99)

| Benchmark | Llama 3-70B | Qwen3 32B |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 4.3% | 66.9% |
| LMArena Math | 1218 | 1399 |
| MATH Level 5 | 22.6% | — |

## Knowledge

- Llama 3-70B: 20.8 (#277)
- Qwen3 32B: 40.0 (#125)

| Benchmark | Llama 3-70B | Qwen3 32B |
|---|---|---|
| GPQA Diamond | 40.6% | 65.7% |
| LMArena Expert | 1149 | 1362 |
| Vectara Hallucination Rate | — | 5.9% |
| MMLU | 79.3% | — |

## Multilingual

- Llama 3-70B: 33.6 (#251)
- Qwen3 32B: 45.6 (#167)

| Benchmark | Llama 3-70B | Qwen3 32B |
|---|---|---|
| LMArena Non-English | 1142 | 1317 |
| LMArena Chinese | 1114 | 1357 |
| LMArena German | 1169 | 1341 |
| LMArena Russian | 1159 | 1311 |
| LMArena French | 1232 | — |
| LMArena Japanese | 1017 | — |
| LMArena Korean | 1017 | — |
| LMArena Spanish | 1241 | — |

## Instruction Following

- Llama 3-70B: 62.5 (#238)
- Qwen3 32B: 68.9 (#179)

| Benchmark | Llama 3-70B | Qwen3 32B |
|---|---|---|
| LMArena Instruction Following | 1194 | 1305 |

## Long Context

- Llama 3-70B: 35.6 (#240)
- Qwen3 32B: 43.8 (#87)

| Benchmark | Llama 3-70B | Qwen3 32B |
|---|---|---|
| LMArena Longer Query | 1174 | 1327 |
| Fiction.LiveBench | — | 74.2% |

## Writing & Preference

- Llama 3-70B: 42.8 (#231)
- Qwen3 32B: 52.9 (#163)

| Benchmark | Llama 3-70B | Qwen3 32B |
|---|---|---|
| LMArena Text | 1221 | 1340 |
| LMArena Creative Writing | 1210 | 1297 |
| LMArena Multi-Turn | 1223 | 1331 |

## FAQ

### Is Llama 3-70B better than Qwen3 32B?

Qwen3 32B is the stronger model overall, scoring 39.2 to 28.8 on the Noometry Index.

### Is Llama 3-70B or Qwen3 32B better for coding?

Qwen3 32B scores higher on coding benchmarks: 37.7 versus 35.8 in the Noometry coding category.

### How many benchmarks do Llama 3-70B and Qwen3 32B share?

18 benchmarks have published results for both models. Llama 3-70B has 31 scored results on Noometry and Qwen3 32B has 26.