Model comparison
Mistral Large 4 vs Qwen2.5 72B Instruct
Mistral Large 4 is the stronger model overall, scoring 43.1 to 31.9 on the Noometry Index.
Last verified . 12 shared benchmarks.
Mistral AI
Alibaba (Qwen)Summary
- They share 12 benchmarks with published results for both. Mistral Large 4 scores higher in 8 categories and Qwen2.5 72B Instruct in 0 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in math, where Mistral Large 4 leads 40.4 to 19.3.
- Mistral Large 4 is cheaper at $0.68 / $2.09 per million input/output tokens, against $1.40 / $5.60 for Qwen2.5 72B Instruct.
- Mistral Large 4 accepts more context: 1.05M tokens versus 131K.
- Qwen2.5 72B Instruct has downloadable open weights; the other is API-only.
Side by side
| Mistral Large 4 | Qwen2.5 72B Instruct | |
|---|---|---|
| Provider | Mistral AI | Alibaba (Qwen) |
| Noometry Index | 43.1 | 31.9 |
| Released | 2026-10-06 | 2024-09 |
| Weights | Proprietary | Open |
| Context window | 1.05M | 131K |
| Max output | 262K | 8K |
| Input $ / M tokens | $0.68 | $1.40 |
| Output $ / M tokens | $2.09 | $5.60 |
| Results tracked | 15 | 43 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Mistral Large 4 leads
Mistral Large 4: 48.6 (#57), Qwen2.5 72B Instruct: 33.2 (#260)
| Benchmark | Mistral Large 4 | Qwen2.5 72B Instruct |
|---|---|---|
| LMArena Coding | 1475 | 1292 |
| LMArena WebDev | 1541 | — |
| WeirdML | — | 16% |
| BigCodeBench Instruct | — | 45.8% |
| BigCodeBench Complete | — | 55.9% |
Agentic & Tool Use Not comparable
Mistral Large 4: —, Qwen2.5 72B Instruct: 22.1 (#133)
| Benchmark | Mistral Large 4 | Qwen2.5 72B Instruct |
|---|---|---|
| TheAgentCompany | — | 5.7% |
| BALROG | — | 16.2% |
| METR Time Horizons | — | 35.8% |
Reasoning Too close to call
Mistral Large 4: 22.5 (#192), Qwen2.5 72B Instruct: 22.3 (#199)
| Benchmark | Mistral Large 4 | Qwen2.5 72B Instruct |
|---|---|---|
| LMArena Hard Prompts | 1444 | 1271 |
| NYT Connections (extended) | 27.4% | — |
| DTBench | — | 62.9% |
| LMCA | — | 13.4% |
| BIG-Bench Hard | — | 79.8% |
| Epoch Capabilities Index | — | 129 |
| ForecastBench | — | 57.5 |
| HellaSwag | — | 84.8% |
| PIQA | — | 82.6% |
| WinoGrande | — | 82.3% |
Math Mistral Large 4 leads
Mistral Large 4: 40.4 (#91), Qwen2.5 72B Instruct: 19.3 (#287)
| Benchmark | Mistral Large 4 | Qwen2.5 72B Instruct |
|---|---|---|
| LMArena Math | 1488 | 1283 |
| OTIS Mock AIME 2024-2025 | — | 8.1% |
| Omni-MATH | — | 33% |
| MATH Level 5 | — | 63.2% |
Knowledge Mistral Large 4 leads
Mistral Large 4: 36.6 (#166), Qwen2.5 72B Instruct: 27.0 (#253)
| Benchmark | Mistral Large 4 | Qwen2.5 72B Instruct |
|---|---|---|
| LMArena Expert | 1447 | 1245 |
| GPQA Diamond | — | 49.1% |
| SimpleQA Verified | 20% | — |
| MMLU-Pro | — | 63.1% |
| Confabulations | — | 19.1% |
| GPQA (HELM) | — | 42.6% |
| ARC (AI2) Challenge | — | 94.5% |
| MMLU | — | 85.3% |
| TriviaQA | — | 71.9% |
Multilingual Mistral Large 4 leads
Mistral Large 4: 52.6 (#82), Qwen2.5 72B Instruct: 41.0 (#213)
| Benchmark | Mistral Large 4 | Qwen2.5 72B Instruct |
|---|---|---|
| LMArena Non-English | 1415 | 1252 |
| LMArena Chinese | 1491 | 1272 |
| LMArena Russian | 1414 | 1264 |
| LMArena French | — | 1280 |
| LMArena German | — | 1234 |
| LMArena Japanese | — | 1180 |
| LMArena Korean | — | 1188 |
| LMArena Spanish | — | 1256 |
Instruction Following Mistral Large 4 leads
Mistral Large 4: 75.0 (#76), Qwen2.5 72B Instruct: 65.5 (#221)
| Benchmark | Mistral Large 4 | Qwen2.5 72B Instruct |
|---|---|---|
| LMArena Instruction Following | 1424 | 1254 |
| IFEval | — | 80.6% |
Long Context Mistral Large 4 leads
Mistral Large 4: 43.6 (#89), Qwen2.5 72B Instruct: 38.9 (#188)
| Benchmark | Mistral Large 4 | Qwen2.5 72B Instruct |
|---|---|---|
| LMArena Longer Query | 1429 | 1282 |
Writing & Preference Mistral Large 4 leads
Mistral Large 4: 60.4 (#97), Qwen2.5 72B Instruct: 46.7 (#215)
| Benchmark | Mistral Large 4 | Qwen2.5 72B Instruct |
|---|---|---|
| LMArena Text | 1427 | 1269 |
| LMArena Creative Writing | 1361 | 1221 |
| LMArena Multi-Turn | 1424 | 1272 |
| WildBench | — | 80.2% |
Frequently asked questions
Is Mistral Large 4 better than Qwen2.5 72B Instruct?
Mistral Large 4 is the stronger model overall, scoring 43.1 to 31.9 on the Noometry Index.
Which is cheaper, Mistral Large 4 or Qwen2.5 72B Instruct?
Mistral Large 4 is cheaper. It lists at $0.68 per million input tokens and $2.09 per million output tokens; Qwen2.5 72B Instruct lists at $1.40 and $5.60.
Is Mistral Large 4 or Qwen2.5 72B Instruct better for coding?
Mistral Large 4 scores higher on coding benchmarks: 48.6 versus 33.2 in the Noometry coding category.
Which has the bigger context window?
Mistral Large 4 does, with 1.05M tokens against 131K.
How many benchmarks do Mistral Large 4 and Qwen2.5 72B Instruct share?
12 benchmarks have published results for both models. Mistral Large 4 has 15 scored results on Noometry and Qwen2.5 72B Instruct has 43.