Model comparison
Mistral Large 3 vs Qwen3.8 27B
Qwen3.8 27B is the stronger model overall, scoring 45.3 to 38.6 on the Noometry Index. Mistral Large 3 costs 3.0× less per token, which makes it the better buy when Qwen3.8 27B's lead doesn't matter for your workload.
Last verified . 21 shared benchmarks.
Mistral AI
Alibaba (Qwen)Summary
- They share 21 benchmarks with published results for both. Mistral Large 3 scores higher in 1 category and Qwen3.8 27B in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Qwen3.8 27B leads 31.5 to 14.6.
- The biggest single-benchmark swing is NYT Connections (extended): 7.5% for Mistral Large 3 and 54.5% for Qwen3.8 27B.
- Mistral Large 3 is cheaper at $0.25 / $0.75 per million input/output tokens, against $0.99 / $1.49 for Qwen3.8 27B.
Side by side
| Mistral Large 3 | Qwen3.8 27B | |
|---|---|---|
| Provider | Mistral AI | Alibaba (Qwen) |
| Noometry Index | 38.6 | 45.3 |
| Released | 2025-12-02 | 2026-08-14 |
| Weights | Open | Open |
| Context window | 262K | 262K |
| Max output | 8K | 33K |
| Input $ / M tokens | $0.25 | $0.99 |
| Output $ / M tokens | $0.75 | $1.49 |
| Results tracked | 24 | 28 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Qwen3.8 27B leads
Mistral Large 3: 34.1 (#230), Qwen3.8 27B: 50.3 (#42)
| Benchmark | Mistral Large 3 | Qwen3.8 27B |
|---|---|---|
| LMArena WebDev | 1230 | 1592 |
| LMArena Coding | 1447 | 1487 |
| SciCode | — | 46.6% |
Reasoning Qwen3.8 27B leads
Mistral Large 3: 14.6 (#314), Qwen3.8 27B: 31.5 (#71)
| Benchmark | Mistral Large 3 | Qwen3.8 27B |
|---|---|---|
| NYT Connections (extended) | 7.5% | 54.5% |
| LMArena Hard Prompts | 1429 | 1459 |
| Kagi LLM Benchmark | 50.9% | — |
| CritPt | — | 5.4% |
| Thematic Generalization | 23% | — |
| DTBench | — | 88% |
| LMCA | — | 41.4% |
| Surface Evolver Bench | — | 45% |
| Epoch Capabilities Index | — | 149.38 |
Math Too close to call
Mistral Large 3: 36.0 (#131), Qwen3.8 27B: 35.7 (#139)
| Benchmark | Mistral Large 3 | Qwen3.8 27B |
|---|---|---|
| LMArena Math | 1415 | 1453 |
| ProofBench | — | 16% |
Knowledge Qwen3.8 27B leads
Mistral Large 3: 35.5 (#173), Qwen3.8 27B: 40.9 (#105)
| Benchmark | Mistral Large 3 | Qwen3.8 27B |
|---|---|---|
| LMArena Expert | 1421 | 1482 |
| Vectara Hallucination Rate | 14.5% | — |
Multimodal Qwen3.8 27B leads
Mistral Large 3: 37.6 (#65), Qwen3.8 27B: 40.5 (#36)
| Benchmark | Mistral Large 3 | Qwen3.8 27B |
|---|---|---|
| LMArena Vision | 1222 | 1270 |
Multilingual Qwen3.8 27B leads
Mistral Large 3: 52.6 (#79), Qwen3.8 27B: 54.0 (#54)
| Benchmark | Mistral Large 3 | Qwen3.8 27B |
|---|---|---|
| LMArena Non-English | 1412 | 1431 |
| LMArena Chinese | 1448 | 1504 |
| LMArena French | 1456 | 1468 |
| LMArena German | 1438 | 1435 |
| LMArena Japanese | 1399 | 1392 |
| LMArena Korean | 1384 | 1389 |
| LMArena Russian | 1412 | 1421 |
| LMArena Spanish | 1437 | 1444 |
Instruction Following Qwen3.8 27B leads
Mistral Large 3: 74.1 (#104), Qwen3.8 27B: 75.8 (#54)
| Benchmark | Mistral Large 3 | Qwen3.8 27B |
|---|---|---|
| LMArena Instruction Following | 1402 | 1436 |
Long Context Qwen3.8 27B leads
Mistral Large 3: 43.1 (#102), Qwen3.8 27B: 44.3 (#66)
| Benchmark | Mistral Large 3 | Qwen3.8 27B |
|---|---|---|
| LMArena Longer Query | 1413 | 1448 |
Writing & Preference Qwen3.8 27B leads
Mistral Large 3: 60.2 (#100), Qwen3.8 27B: 65.9 (#41)
| Benchmark | Mistral Large 3 | Qwen3.8 27B |
|---|---|---|
| LMArena Text | 1427 | 1441 |
| LMArena Creative Writing | 1385 | 1382 |
| EQ-Bench Creative Writing | 1412 | 1671 |
| LMArena Multi-Turn | 1429 | 1443 |
Frequently asked questions
Is Mistral Large 3 better than Qwen3.8 27B?
Qwen3.8 27B is the stronger model overall, scoring 45.3 to 38.6 on the Noometry Index. Mistral Large 3 costs 3.0× less per token, which makes it the better buy when Qwen3.8 27B's lead doesn't matter for your workload.
Which is cheaper, Mistral Large 3 or Qwen3.8 27B?
Mistral Large 3 is cheaper. It lists at $0.25 per million input tokens and $0.75 per million output tokens; Qwen3.8 27B lists at $0.99 and $1.49.
Is Mistral Large 3 or Qwen3.8 27B better for coding?
Qwen3.8 27B scores higher on coding benchmarks: 50.3 versus 34.1 in the Noometry coding category.
Which has the bigger context window?
Both accept 262K tokens.
How many benchmarks do Mistral Large 3 and Qwen3.8 27B share?
21 benchmarks have published results for both models. Mistral Large 3 has 24 scored results on Noometry and Qwen3.8 27B has 28.