Model comparison
Mistral Large vs Qwen3.5 27B
Qwen3.5 27B is the stronger model overall, scoring 41.9 to 31.9 on the Noometry Index.
Last verified . 21 shared benchmarks.
Mistral AI
Alibaba (Qwen)Summary
- They share 21 benchmarks with published results for both. Mistral Large scores higher in 0 categories and Qwen3.5 27B in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where Qwen3.5 27B leads 38.8 to 18.2.
- The biggest single-benchmark swing is DTBench: 65.1% for Mistral Large and 82.4% for Qwen3.5 27B.
- Qwen3.5 27B is cheaper at $0.30 / $2.40 per million input/output tokens, against $2 / $6 for Mistral Large.
- Qwen3.5 27B accepts more context: 262K tokens versus 131K.
Side by side
| Mistral Large | Qwen3.5 27B | |
|---|---|---|
| Provider | Mistral AI | Alibaba (Qwen) |
| Noometry Index | 31.9 | 41.9 |
| Released | 2024-02-26 | 2026-02-23 |
| Weights | Open | Open |
| Context window | 131K | 262K |
| Max output | 16K | 66K |
| Input $ / M tokens | $2 | $0.30 |
| Output $ / M tokens | $6 | $2.40 |
| Results tracked | 51 | 28 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Qwen3.5 27B leads
Mistral Large: 34.3 (#240), Qwen3.5 27B: 38.9 (#168)
| Benchmark | Mistral Large | Qwen3.5 27B |
|---|---|---|
| LMArena Coding | 1277 | 1427 |
| ALE-Bench | 264.7 | 349.45 |
| LMArena WebDev | — | 1358 |
| SciCode | 36.2% | — |
| WeirdML | — | 39.5% |
| BigCodeBench Instruct | 30% | — |
| LiveBench Coding | 47.1% | — |
| BigCodeBench Complete | 38.3% | — |
| HumanEval+ | 62.2% | — |
| MBPP+ | 59.5% | — |
Agentic & Tool Use Not comparable
Mistral Large: 28.6 (#89), Qwen3.5 27B: —
| Benchmark | Mistral Large | Qwen3.5 27B |
|---|---|---|
| Berkeley Function Calling Leaderboard | 38.4% | — |
| Vending-Bench 2 | — | 201.98 |
Reasoning Qwen3.5 27B leads
Mistral Large: 15.8 (#310), Qwen3.5 27B: 27.5 (#117)
| Benchmark | Mistral Large | Qwen3.5 27B |
|---|---|---|
| LMArena Hard Prompts | 1257 | 1414 |
| DTBench | 65.1% | 82.4% |
| LMCA | 16.7% | 34% |
| SimpleBench | 22.5% | — |
| NYT Connections (extended) | — | 47.9% |
| CritPt | 0% | — |
| Thematic Generalization | — | 45.5% |
| LiveBench Reasoning | 43.5% | — |
| LiveBench Data Analysis | 50.1% | — |
| Epoch Capabilities Index | 128.52 | — |
| ForecastBench | 57.1 | — |
| LiveBench | 48.4% | — |
Math Qwen3.5 27B leads
Mistral Large: 18.2 (#291), Qwen3.5 27B: 38.8 (#127)
| Benchmark | Mistral Large | Qwen3.5 27B |
|---|---|---|
| LMArena Math | 1262 | 1429 |
| MathArena Final-Answer Competitions | — | 56.7% |
| OTIS Mock AIME 2024-2025 | 8.5% | — |
| Omni-MATH | 28.1% | — |
| LiveBench Math | 42.5% | — |
| MATH Level 5 | 50.3% | — |
| FrontierMath (Feb 2025 set) | 0.3% | — |
Knowledge Qwen3.5 27B leads
Mistral Large: 30.1 (#230), Qwen3.5 27B: 38.0 (#150)
| Benchmark | Mistral Large | Qwen3.5 27B |
|---|---|---|
| Vectara Hallucination Rate | 4.5% | 12.1% |
| LMArena Expert | 1232 | 1428 |
| GPQA Diamond | 51.3% | — |
| MMLU-Pro | 59.9% | — |
| Confabulations | 21.4% | — |
| GPQA (HELM) | 43.5% | — |
| MMLU | 80% | — |
Multimodal Not comparable
Mistral Large: —, Qwen3.5 27B: 39.4 (#59)
| Benchmark | Mistral Large | Qwen3.5 27B |
|---|---|---|
| LMArena Vision | — | 1241 |
Multilingual Qwen3.5 27B leads
Mistral Large: 40.0 (#219), Qwen3.5 27B: 50.8 (#115)
| Benchmark | Mistral Large | Qwen3.5 27B |
|---|---|---|
| LMArena Non-English | 1237 | 1390 |
| LMArena Chinese | 1240 | 1478 |
| LMArena French | 1325 | 1410 |
| LMArena German | 1254 | 1393 |
| LMArena Japanese | 1188 | 1345 |
| LMArena Korean | 1202 | 1358 |
| LMArena Russian | 1257 | 1390 |
| LMArena Spanish | 1268 | 1407 |
Instruction Following Qwen3.5 27B leads
Mistral Large: 67.9 (#191), Qwen3.5 27B: 73.5 (#119)
| Benchmark | Mistral Large | Qwen3.5 27B |
|---|---|---|
| LMArena Instruction Following | 1249 | 1393 |
| LiveBench Instruction Following | 67.9% | — |
| IFEval | 87.7% | — |
Long Context Qwen3.5 27B leads
Mistral Large: 38.3 (#199), Qwen3.5 27B: 43.1 (#106)
| Benchmark | Mistral Large | Qwen3.5 27B |
|---|---|---|
| LMArena Longer Query | 1261 | 1413 |
Writing & Preference Qwen3.5 27B leads
Mistral Large: 40.7 (#242), Qwen3.5 27B: 59.3 (#111)
| Benchmark | Mistral Large | Qwen3.5 27B |
|---|---|---|
| LMArena Text | 1266 | 1409 |
| LMArena Creative Writing | 1243 | 1362 |
| LMArena Multi-Turn | 1260 | 1410 |
| Short-Story Creative Writing | 69% | — |
| EQ-Bench Creative Writing | 985 | — |
| WildBench | 80.1% | — |
| LiveBench Language | 39.4% | — |
Frequently asked questions
Is Mistral Large better than Qwen3.5 27B?
Qwen3.5 27B is the stronger model overall, scoring 41.9 to 31.9 on the Noometry Index.
Which is cheaper, Mistral Large or Qwen3.5 27B?
Qwen3.5 27B is cheaper. It lists at $0.30 per million input tokens and $2.40 per million output tokens; Mistral Large lists at $2 and $6.
Is Mistral Large or Qwen3.5 27B better for coding?
Qwen3.5 27B scores higher on coding benchmarks: 38.9 versus 34.3 in the Noometry coding category.
Which has the bigger context window?
Qwen3.5 27B does, with 262K tokens against 131K.
How many benchmarks do Mistral Large and Qwen3.5 27B share?
21 benchmarks have published results for both models. Mistral Large has 51 scored results on Noometry and Qwen3.5 27B has 28.