Model comparison
Llama 3.1-405B vs Qwen3.8 27B
Qwen3.8 27B is the stronger model overall, scoring 46.0 to 30.7 on the Noometry Index.
Last verified . 20 shared benchmarks.
Meta
Alibaba (Qwen)Summary
- They share 20 benchmarks with published results for both. Llama 3.1-405B scores higher in 0 categories and Qwen3.8 27B in 9 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where Qwen3.8 27B leads 65.8 to 38.9.
- The biggest single-benchmark swing is DTBench: 61.4% for Llama 3.1-405B and 88% for Qwen3.8 27B.
Side by side
| Llama 3.1-405B | Qwen3.8 27B | |
|---|---|---|
| Provider | Meta | Alibaba (Qwen) |
| Noometry Index | 30.7 | 46.0 |
| Released | 2024-07-23 | 2026-08-14 |
| Weights | Open | Open |
| Context window | — | 262K |
| Max output | — | 33K |
| Input $ / M tokens | — | $0.99 |
| Output $ / M tokens | — | $1.49 |
| Results tracked | 42 | 31 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Qwen3.8 27B leads
Llama 3.1-405B: 33.1 (#262), Qwen3.8 27B: 50.5 (#44)
| Benchmark | Llama 3.1-405B | Qwen3.8 27B |
|---|---|---|
| LMArena Coding | 1291 | 1482 |
| LMArena WebDev | — | 1593 |
| SciCode | — | 46.6% |
| WeirdML | 21.4% | — |
Agentic & Tool Use Qwen3.8 27B leads
Llama 3.1-405B: 21.0 (#140), Qwen3.8 27B: 32.9 (#57)
| Benchmark | Llama 3.1-405B | Qwen3.8 27B |
|---|---|---|
| APEX-Agents | — | 47.5% |
| TheAgentCompany | 7.4% | — |
| Cybench | 7.5% | — |
Reasoning Qwen3.8 27B leads
Llama 3.1-405B: 16.8 (#300), Qwen3.8 27B: 41.0 (#54)
| Benchmark | Llama 3.1-405B | Qwen3.8 27B |
|---|---|---|
| LMArena Hard Prompts | 1269 | 1460 |
| DTBench | 61.4% | 88% |
| Epoch Capabilities Index | 128.75 | 149.38 |
| ARC-AGI-2 | — | 42.4% |
| SimpleBench | 23% | — |
| Kagi LLM Benchmark | 45% | — |
| NYT Connections (extended) | — | 54.5% |
| ARC-AGI-1 | — | 87.5% |
| CritPt | — | 5.4% |
| LMCA | — | 41.4% |
| Surface Evolver Bench | — | 45% |
| BIG-Bench Hard | 82.9% | — |
| ForecastBench | 59.9 | — |
| HellaSwag | 89.2% | — |
| PIQA | 85.9% | — |
| WinoGrande | 89.2% | — |
Math Qwen3.8 27B leads
Llama 3.1-405B: 18.4 (#290), Qwen3.8 27B: 37.1 (#161)
| Benchmark | Llama 3.1-405B | Qwen3.8 27B |
|---|---|---|
| LMArena Math | 1281 | 1456 |
| OTIS Mock AIME 2024-2025 | 9.7% | — |
| ProofBench | — | 16% |
| Omni-MATH | 24.9% | — |
| MATH Level 5 | 49.8% | — |
Knowledge Qwen3.8 27B leads
Llama 3.1-405B: 30.4 (#227), Qwen3.8 27B: 41.6 (#109)
| Benchmark | Llama 3.1-405B | Qwen3.8 27B |
|---|---|---|
| LMArena Expert | 1243 | 1482 |
| GPQA Diamond | 50.9% | — |
| MMLU-Pro | 72.3% | — |
| Confabulations | 17.6% | — |
| GPQA (HELM) | 52.2% | — |
| ARC (AI2) Challenge | 95.3% | — |
| MMLU | 84.5% | — |
| TriviaQA | 82.7% | — |
Multimodal Not comparable
Llama 3.1-405B: —, Qwen3.8 27B: 41.3 (#37)
| Benchmark | Llama 3.1-405B | Qwen3.8 27B |
|---|---|---|
| LMArena Vision | — | 1271 |
Multilingual Qwen3.8 27B leads
Llama 3.1-405B: 40.7 (#214), Qwen3.8 27B: 53.7 (#60)
| Benchmark | Llama 3.1-405B | Qwen3.8 27B |
|---|---|---|
| LMArena Non-English | 1248 | 1430 |
| LMArena Chinese | 1242 | 1504 |
| LMArena French | 1279 | 1465 |
| LMArena German | 1252 | 1438 |
| LMArena Japanese | 1208 | 1384 |
| LMArena Korean | 1184 | 1393 |
| LMArena Russian | 1265 | 1415 |
| LMArena Spanish | 1260 | 1448 |
Instruction Following Qwen3.8 27B leads
Llama 3.1-405B: 65.9 (#214), Qwen3.8 27B: 75.8 (#53)
| Benchmark | Llama 3.1-405B | Qwen3.8 27B |
|---|---|---|
| LMArena Instruction Following | 1259 | 1439 |
| IFEval | 81.1% | — |
Long Context Qwen3.8 27B leads
Llama 3.1-405B: 38.4 (#197), Qwen3.8 27B: 44.3 (#70)
| Benchmark | Llama 3.1-405B | Qwen3.8 27B |
|---|---|---|
| LMArena Longer Query | 1266 | 1450 |
Writing & Preference Qwen3.8 27B leads
Llama 3.1-405B: 38.9 (#251), Qwen3.8 27B: 65.8 (#43)
| Benchmark | Llama 3.1-405B | Qwen3.8 27B |
|---|---|---|
| LMArena Text | 1284 | 1441 |
| LMArena Creative Writing | 1262 | 1384 |
| EQ-Bench Creative Writing | 870 | 1671 |
| LMArena Multi-Turn | 1297 | 1441 |
| WildBench | 78.3% | — |
Frequently asked questions
Is Llama 3.1-405B better than Qwen3.8 27B?
Qwen3.8 27B is the stronger model overall, scoring 46.0 to 30.7 on the Noometry Index.
Is Llama 3.1-405B or Qwen3.8 27B better for coding?
Qwen3.8 27B scores higher on coding benchmarks: 50.5 versus 33.1 in the Noometry coding category.
How many benchmarks do Llama 3.1-405B and Qwen3.8 27B share?
20 benchmarks have published results for both models. Llama 3.1-405B has 42 scored results on Noometry and Qwen3.8 27B has 31.