Model comparison
Mistral Large 3 vs Qwen3.5 27B
Qwen3.5 27B is the stronger model overall, scoring 41.9 to 39.1 on the Noometry Index. Mistral Large 3 costs 2.2× less per token, which makes it the better buy when Qwen3.5 27B's lead doesn't matter for your workload.
Last verified . 22 shared benchmarks.
Mistral AI
Alibaba (Qwen)Summary
- They share 22 benchmarks with published results for both. Mistral Large 3 scores higher in 4 categories and Qwen3.5 27B in 5 categories; 5 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Qwen3.5 27B leads 27.5 to 15.2.
- The biggest single-benchmark swing is NYT Connections (extended): 7.5% for Mistral Large 3 and 47.9% for Qwen3.5 27B.
- Mistral Large 3 is cheaper at $0.25 / $0.75 per million input/output tokens, against $0.30 / $2.40 for Qwen3.5 27B.
Side by side
| Mistral Large 3 | Qwen3.5 27B | |
|---|---|---|
| Provider | Mistral AI | Alibaba (Qwen) |
| Noometry Index | 39.1 | 41.9 |
| Released | 2025-12-02 | 2026-02-23 |
| Weights | Open | Open |
| Context window | 262K | 262K |
| Max output | 8K | 66K |
| Input $ / M tokens | $0.25 | $0.30 |
| Output $ / M tokens | $0.75 | $2.40 |
| Results tracked | 24 | 28 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Qwen3.5 27B leads
Mistral Large 3: 34.4 (#237), Qwen3.5 27B: 38.9 (#168)
| Benchmark | Mistral Large 3 | Qwen3.5 27B |
|---|---|---|
| LMArena WebDev | 1230 | 1358 |
| LMArena Coding | 1448 | 1427 |
| WeirdML | — | 39.5% |
| ALE-Bench | — | 349.45 |
Agentic & Tool Use Not comparable
Mistral Large 3: —, Qwen3.5 27B: —
| Benchmark | Mistral Large 3 | Qwen3.5 27B |
|---|---|---|
| Vending-Bench 2 | — | 201.98 |
Reasoning Qwen3.5 27B leads
Mistral Large 3: 15.2 (#319), Qwen3.5 27B: 27.5 (#117)
| Benchmark | Mistral Large 3 | Qwen3.5 27B |
|---|---|---|
| NYT Connections (extended) | 7.5% | 47.9% |
| Thematic Generalization | 23% | 45.5% |
| LMArena Hard Prompts | 1429 | 1414 |
| Kagi LLM Benchmark | 50.9% | — |
| DTBench | — | 82.4% |
| LMCA | — | 34% |
Math Too close to call
Mistral Large 3: 38.7 (#129), Qwen3.5 27B: 38.8 (#127)
| Benchmark | Mistral Large 3 | Qwen3.5 27B |
|---|---|---|
| LMArena Math | 1414 | 1429 |
| MathArena Final-Answer Competitions | — | 56.7% |
Knowledge Qwen3.5 27B leads
Mistral Large 3: 36.0 (#177), Qwen3.5 27B: 38.0 (#150)
| Benchmark | Mistral Large 3 | Qwen3.5 27B |
|---|---|---|
| Vectara Hallucination Rate | 14.5% | 12.1% |
| LMArena Expert | 1421 | 1428 |
Multimodal Qwen3.5 27B leads
Mistral Large 3: 38.2 (#66), Qwen3.5 27B: 39.4 (#59)
| Benchmark | Mistral Large 3 | Qwen3.5 27B |
|---|---|---|
| LMArena Vision | 1221 | 1241 |
Multilingual Mistral Large 3 leads
Mistral Large 3: 52.5 (#84), Qwen3.5 27B: 50.8 (#115)
| Benchmark | Mistral Large 3 | Qwen3.5 27B |
|---|---|---|
| LMArena Non-English | 1413 | 1390 |
| LMArena Chinese | 1447 | 1478 |
| LMArena French | 1455 | 1410 |
| LMArena German | 1437 | 1393 |
| LMArena Japanese | 1394 | 1345 |
| LMArena Korean | 1384 | 1358 |
| LMArena Russian | 1411 | 1390 |
| LMArena Spanish | 1440 | 1407 |
Instruction Following Too close to call
Mistral Large 3: 74.0 (#108), Qwen3.5 27B: 73.5 (#119)
| Benchmark | Mistral Large 3 | Qwen3.5 27B |
|---|---|---|
| LMArena Instruction Following | 1403 | 1393 |
Long Context Too close to call
Mistral Large 3: 43.1 (#105), Qwen3.5 27B: 43.1 (#106)
| Benchmark | Mistral Large 3 | Qwen3.5 27B |
|---|---|---|
| LMArena Longer Query | 1413 | 1413 |
Writing & Preference Too close to call
Mistral Large 3: 60.0 (#101), Qwen3.5 27B: 59.3 (#111)
| Benchmark | Mistral Large 3 | Qwen3.5 27B |
|---|---|---|
| LMArena Text | 1428 | 1409 |
| LMArena Creative Writing | 1386 | 1362 |
| LMArena Multi-Turn | 1429 | 1410 |
| EQ-Bench Creative Writing | 1412 | — |
Frequently asked questions
Is Mistral Large 3 better than Qwen3.5 27B?
Qwen3.5 27B is the stronger model overall, scoring 41.9 to 39.1 on the Noometry Index. Mistral Large 3 costs 2.2× less per token, which makes it the better buy when Qwen3.5 27B's lead doesn't matter for your workload.
Which is cheaper, Mistral Large 3 or Qwen3.5 27B?
Mistral Large 3 is cheaper. It lists at $0.25 per million input tokens and $0.75 per million output tokens; Qwen3.5 27B lists at $0.30 and $2.40.
Is Mistral Large 3 or Qwen3.5 27B better for coding?
Qwen3.5 27B scores higher on coding benchmarks: 38.9 versus 34.4 in the Noometry coding category.
Which has the bigger context window?
Both accept 262K tokens.
How many benchmarks do Mistral Large 3 and Qwen3.5 27B share?
22 benchmarks have published results for both models. Mistral Large 3 has 24 scored results on Noometry and Qwen3.5 27B has 28.