Model comparison
Llama 3-8B vs Qwen3 32B
Qwen3 32B is the stronger model overall, scoring 39.2 to 25.5 on the Noometry Index.
Last verified . 18 shared benchmarks.
Meta
Alibaba (Qwen)Summary
- They share 18 benchmarks with published results for both. Llama 3-8B scores higher in 0 categories and Qwen3 32B in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where Qwen3 32B leads 40.0 to 7.8.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 1.9% for Llama 3-8B and 66.9% for Qwen3 32B.
Side by side
| Llama 3-8B | Qwen3 32B | |
|---|---|---|
| Provider | Meta | Alibaba (Qwen) |
| Noometry Index | 25.5 | 39.2 |
| Released | 2024-04-18 | 2025-04 |
| Weights | Open | Open |
| Context window | — | 131K |
| Max output | — | 16K |
| Input $ / M tokens | — | $0.70 |
| Output $ / M tokens | — | $2.80 |
| Results tracked | 34 | 26 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Qwen3 32B leads
Llama 3-8B: 31.0 (#289), Qwen3 32B: 37.7 (#190)
| Benchmark | Llama 3-8B | Qwen3 32B |
|---|---|---|
| LMArena Coding | 1152 | 1358 |
| Aider Polyglot | — | 40% |
| SciCode | — | 35.4% |
| BigCodeBench Instruct | 31.9% | — |
| BigCodeBench Complete | 36.9% | — |
| HumanEval+ | 56.7% | — |
| MBPP+ | 54.8% | — |
Agentic & Tool Use Not comparable
Llama 3-8B: —, Qwen3 32B: 32.6 (#62)
| Benchmark | Llama 3-8B | Qwen3 32B |
|---|---|---|
| Berkeley Function Calling Leaderboard | — | 48.7% |
Reasoning Qwen3 32B leads
Llama 3-8B: 14.3 (#326), Qwen3 32B: 20.2 (#241)
| Benchmark | Llama 3-8B | Qwen3 32B |
|---|---|---|
| Chess Puzzles | 0% | 5% |
| LMArena Hard Prompts | 1133 | 1334 |
| DTBench | 43.9% | 67.5% |
| Epoch Capabilities Index | 116.45 | 138.51 |
| Kagi LLM Benchmark | — | 54.9% |
| CritPt | — | 0.3% |
| LMCA | — | 17.3% |
| Adversarial NLI | 57.3% | — |
| ForecastBench | 58.6 | — |
| WinoGrande | 75.7% | — |
Math Qwen3 32B leads
Llama 3-8B: 8.8 (#323), Qwen3 32B: 39.7 (#99)
| Benchmark | Llama 3-8B | Qwen3 32B |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 1.9% | 66.9% |
| LMArena Math | 1151 | 1399 |
| MATH Level 5 | 6.1% | — |
Knowledge Qwen3 32B leads
Llama 3-8B: 7.8 (#308), Qwen3 32B: 40.0 (#125)
| Benchmark | Llama 3-8B | Qwen3 32B |
|---|---|---|
| GPQA Diamond | 26.1% | 65.7% |
| LMArena Expert | 1113 | 1362 |
| Vectara Hallucination Rate | — | 5.9% |
| ARC (AI2) Challenge | 82.8% | — |
| MMLU | 68.8% | — |
| OpenBookQA | 82.6% | — |
| TriviaQA | 67.7% | — |
Multilingual Qwen3 32B leads
Llama 3-8B: 30.8 (#261), Qwen3 32B: 45.6 (#167)
| Benchmark | Llama 3-8B | Qwen3 32B |
|---|---|---|
| LMArena Non-English | 1098 | 1317 |
| LMArena Chinese | 1076 | 1357 |
| LMArena German | 1104 | 1341 |
| LMArena Russian | 1109 | 1311 |
| LMArena French | 1159 | — |
| LMArena Japanese | 967 | — |
| LMArena Korean | 1004 | — |
| LMArena Spanish | 1173 | — |
Instruction Following Qwen3 32B leads
Llama 3-8B: 58.4 (#260), Qwen3 32B: 68.9 (#179)
| Benchmark | Llama 3-8B | Qwen3 32B |
|---|---|---|
| LMArena Instruction Following | 1127 | 1305 |
Long Context Qwen3 32B leads
Llama 3-8B: 34.2 (#251), Qwen3 32B: 43.8 (#87)
| Benchmark | Llama 3-8B | Qwen3 32B |
|---|---|---|
| LMArena Longer Query | 1128 | 1327 |
| Fiction.LiveBench | — | 74.2% |
Writing & Preference Qwen3 32B leads
Llama 3-8B: 37.5 (#256), Qwen3 32B: 52.9 (#163)
| Benchmark | Llama 3-8B | Qwen3 32B |
|---|---|---|
| LMArena Text | 1166 | 1340 |
| LMArena Creative Writing | 1150 | 1297 |
| LMArena Multi-Turn | 1152 | 1331 |
Frequently asked questions
Is Llama 3-8B better than Qwen3 32B?
Qwen3 32B is the stronger model overall, scoring 39.2 to 25.5 on the Noometry Index.
Is Llama 3-8B or Qwen3 32B better for coding?
Qwen3 32B scores higher on coding benchmarks: 37.7 versus 31.0 in the Noometry coding category.
How many benchmarks do Llama 3-8B and Qwen3 32B share?
18 benchmarks have published results for both models. Llama 3-8B has 34 scored results on Noometry and Qwen3 32B has 26.