Mistral Medium 3.1 vs Qwen2.5 72B Instruct
Mistral Medium 3.1 leads Qwen2.5 72B Instruct on the Artificial Analysis Intelligence Index, 9.2 to 7.7, and Qwen2.5 72B Instruct costs 5.0x less per output token ($0.40 vs $2.00 per million). Updated October 21, 2026; 39 benchmarks have a result for both.
Mistral Medium 3.1 or Qwen2.5 72B Instruct: which to pick
Neither model has a coding score yet, so the two cannot be ranked on coding.
Pick Mistral Medium 3.1 for a larger context window (131K against 33K tokens) and work measured by MMLU Pro (Artificial Analysis run), where it leads.
- Shared results
- 39
- Mistral Medium 3.1 only
- 35
- Qwen2.5 72B Instruct only
- 14
- Like-for-like categories
- 7 of 9
Only benchmarks that tested both models count toward a pick; a result only one model has is shown but not compared.
Which one for your work
Each card is decided by one number both models carry. Where only one has it, the card says so and makes no call.
Coding
Vibe Coding Index, six public coding benchmarks
No call
Neither has a coding score yet.
One side not measured
Intelligence
AA Intelligence Index
Mistral Medium 3.1
Mistral Medium 3.1 leads, 9.2 against 7.7.
Same board, both measured
Agentic work
AA Agentic Index
No call
Only Mistral Medium 3.1 has a AA Agentic Index result, 1.9.
Directional only
Math
AIME
No call
Only Mistral Medium 3.1 has a AIME result, 42.3%.
Directional only
Knowledge
MMLU Pro
No call
Only Mistral Medium 3.1 has a MMLU Pro result, 75.3%.
Directional only
Web design
Design Arena: websites
No call
Only Mistral Medium 3.1 has a Design Arena: websites result, 1133.
Directional only
Long documents
Listed context window
Mistral Medium 3.1
Mistral Medium 3.1 takes 131K tokens against 33K.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
Qwen2.5 72B Instruct
Mistral Medium 3.1 $0.0014, Qwen2.5 72B Instruct $0.0006 per turn.
List prices
At a glance
- Coding index
- Mistral Medium 3.1 not benchmarked, Qwen2.5 72B Instruct not benchmarked.
- Widest category
- General, 23.9 points apart on average over 19 shared benchmarks.
- Widest single gap
- LMArena Text: German: Mistral Medium 3.1 1432 against Qwen2.5 72B Instruct 1234, 33.2 points apart on a 0 to 100 scale.
- Output price
- Mistral Medium 3.1 $2.00 against Qwen2.5 72B Instruct $0.40 per million output tokens, 5.0x apart.
- Context window
- Mistral Medium 3.1 takes 131K tokens, Qwen2.5 72B Instruct 33K.
Benchmark by benchmark
First the score: boards each model leads, per category, where both have a result. Then every board, a category to a fold; bars are each result on a 0 to 100 scale, so compare the two in a row, not one row with another.
| Category | Mistral Medium 3.1 | Bars | Qwen2.5 72B Instruct | Ties |
|---|---|---|---|---|
| Coding | 4 | 0 | ||
| Intelligence | 1 | 0 | ||
| Agentic work | 1 | 0 | ||
| Reasoning | 5 | 0 | ||
| Math | 4 | 0 | ||
| Knowledge | 4 | 1 | ||
| General | 19 | 0 | ||
| All shared boards | 38 | 1 |
CodingLike-for-likeSix of these boards feed the Vibe Coding Index40
| Benchmark | Mistral Medium 3.1 | Qwen2.5 72B Instruct | Leader |
|---|---|---|---|
| AA Coding Index | 20.5 | Not measured | Only Mistral Medium 3.1 measured |
| LiveCodeBench (Vals AI) | 44.8% | Not measured | Only Mistral Medium 3.1 measured |
| LiveCodeBench (Artificial Analysis run) | 40.6% | 27.6% | Mistral Medium 3.1 |
| SciCode | 32.4% | Not measured | Only Mistral Medium 3.1 measured |
| Terminal-Bench Hard (Artificial Analysis run) | 10.6% | 4.5% | Mistral Medium 3.1 |
| WeirdML | Not measured | 16.0% | Only Qwen2.5 72B Instruct measured |
| ALE-Bench | 210.2 | Not measured | Only Mistral Medium 3.1 measured |
| LMArena Text: coding | 1433 | 1292 | Mistral Medium 3.1 |
| LMArena Text: software and IT | 1436 | 1292 | Mistral Medium 3.1 |
| Neither model has a result on the other 30 boards here | |||
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index10
| Benchmark | Mistral Medium 3.1 | Qwen2.5 72B Instruct | Leader |
|---|---|---|---|
| AA Intelligence Index | 9.2 | 7.7 | Mistral Medium 3.1 |
| Epoch Capabilities Index | Not measured | 129.0 | Only Qwen2.5 72B Instruct measured |
Agentic workLike-for-likePublic boards, shown as published, kept out of the index10
| Benchmark | Mistral Medium 3.1 | Qwen2.5 72B Instruct | Leader |
|---|---|---|---|
| AA Agentic Index | 1.9 | Not measured | Only Mistral Medium 3.1 measured |
| τ²-Bench Telecom | 40.6% | 34.5% | Mistral Medium 3.1 |
| METR time horizons | Not measured | 35.8% | Only Qwen2.5 72B Instruct measured |
| BALROG | Not measured | 16.2% | Only Qwen2.5 72B Instruct measured |
| TheAgentCompany | Not measured | 5.7% | Only Qwen2.5 72B Instruct measured |
| Neither model has a result on the other 36 boards here | |||
ReasoningLike-for-likePublic boards, shown as published, kept out of the index50
| Benchmark | Mistral Medium 3.1 | Qwen2.5 72B Instruct | Leader |
|---|---|---|---|
| MMMU Pro (Vals AI) | 63.0% | Not measured | Only Mistral Medium 3.1 measured |
| GPQA Diamond (Artificial Analysis run) | 58.8% | 49.1% | Mistral Medium 3.1 |
| Humanity's Last Exam (Artificial Analysis run) | 4.7% | 3.6% | Mistral Medium 3.1 |
| GPQA Diamond (Epoch AI run) | Not measured | 49.1% | Only Qwen2.5 72B Instruct measured |
| BIG-Bench Hard | Not measured | 79.8% | Only Qwen2.5 72B Instruct measured |
| LMArena Text: hard prompts | 1426 | 1271 | Mistral Medium 3.1 |
| LMArena Text: expert | 1408 | 1245 | Mistral Medium 3.1 |
| LMArena Text: hard prompts in English | 1435 | 1279 | Mistral Medium 3.1 |
| LMArena Vision: diagrams | 1175 | Not measured | Only Mistral Medium 3.1 measured |
| LMArena Vision: homework | 1165 | Not measured | Only Mistral Medium 3.1 measured |
| Neither model has a result on the other 12 boards here | |||
MathLike-for-likePublic boards, shown as published, kept out of the index40
| Benchmark | Mistral Medium 3.1 | Qwen2.5 72B Instruct | Leader |
|---|---|---|---|
| AIME | 42.3% | Not measured | Only Mistral Medium 3.1 measured |
| MATH 500 (Vals AI) | 87.0% | Not measured | Only Mistral Medium 3.1 measured |
| MGSM (Vals AI) | 90.9% | Not measured | Only Mistral Medium 3.1 measured |
| AA Math Index | 38.3 | 14.0 | Mistral Medium 3.1 |
| AIME 2025 (Artificial Analysis run) | 38.3% | 14.0% | Mistral Medium 3.1 |
| OTIS Mock AIME | Not measured | 8.1% | Only Qwen2.5 72B Instruct measured |
| MATH level 5 | Not measured | 63.2% | Only Qwen2.5 72B Instruct measured |
| LMArena Text: math | 1408 | 1283 | Mistral Medium 3.1 |
| LMArena Text: mathematical work | 1412 | 1290 | Mistral Medium 3.1 |
| Neither model has a result on the other 4 boards here | |||
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index41
| Benchmark | Mistral Medium 3.1 | Qwen2.5 72B Instruct | Leader |
|---|---|---|---|
| MMLU Pro | 75.3% | Not measured | Only Mistral Medium 3.1 measured |
| LegalBench (Vals AI) | 62.0% | Not measured | Only Mistral Medium 3.1 measured |
| CorpFin (Vals AI) | 50.7% | Not measured | Only Mistral Medium 3.1 measured |
| TaxEval (Vals AI) | 70.3% | Not measured | Only Mistral Medium 3.1 measured |
| MortgageTax (Vals AI) | 36.5% | Not measured | Only Mistral Medium 3.1 measured |
| MedQA (Vals AI) | 78.2% | Not measured | Only Mistral Medium 3.1 measured |
| MMLU Pro (Artificial Analysis run) | 68.3% | 72.0% | Qwen2.5 72B Instruct |
| MMLU | Not measured | 85.3% | Only Qwen2.5 72B Instruct measured |
| AI2 Reasoning Challenge | Not measured | 94.5% | Only Qwen2.5 72B Instruct measured |
| LMArena Text: science | 1444 | 1275 | Mistral Medium 3.1 |
| LMArena Text: medicine and healthcare | 1449 | 1258 | Mistral Medium 3.1 |
| LMArena Text: legal and government | 1439 | 1288 | Mistral Medium 3.1 |
| LMArena Text: business and finance | 1418 | 1253 | Mistral Medium 3.1 |
| Neither model has a result on the other 5 boards here | |||
Web designDirectional onlyPublic boards, shown as published, kept out of the index8 boards, none shared
| Benchmark | Mistral Medium 3.1 | Qwen2.5 72B Instruct | Leader |
|---|---|---|---|
| Design Arena: websites | 1133 | Not measured | Only Mistral Medium 3.1 measured |
| Design Arena: UI components | 1100 | Not measured | Only Mistral Medium 3.1 measured |
| Design Arena: game dev | 1074 | Not measured | Only Mistral Medium 3.1 measured |
| Design Arena: data viz | 1144 | Not measured | Only Mistral Medium 3.1 measured |
| Design Arena: 3D | 1081 | Not measured | Only Mistral Medium 3.1 measured |
| Design Arena: all code categories | 1120 | Not measured | Only Mistral Medium 3.1 measured |
| Design Arena: SVG | 993 | Not measured | Only Mistral Medium 3.1 measured |
| Design Arena: ASCII art | 997 | Not measured | Only Mistral Medium 3.1 measured |
| Neither model has a result on the other 1 board here | |||
GeneralLike-for-likePublic boards, shown as published, kept out of the index190
| Benchmark | Mistral Medium 3.1 | Qwen2.5 72B Instruct | Leader |
|---|---|---|---|
| IFBench | 39.8% | 36.9% | Mistral Medium 3.1 |
| Long context reasoning (AA-LCR) | 42.7% | Not measured | Only Mistral Medium 3.1 measured |
| DTBench | 67.5% | 62.9% | Mistral Medium 3.1 |
| LMCA | 19.1% | 13.4% | Mistral Medium 3.1 |
| HellaSwag | Not measured | 84.8% | Only Qwen2.5 72B Instruct measured |
| WinoGrande | Not measured | 82.3% | Only Qwen2.5 72B Instruct measured |
| PIQA | Not measured | 82.6% | Only Qwen2.5 72B Instruct measured |
| LMArena Text | 1424 | 1269 | Mistral Medium 3.1 |
| LMArena Text: instruction following | 1398 | 1254 | Mistral Medium 3.1 |
| LMArena Text: creative writing | 1391 | 1221 | Mistral Medium 3.1 |
| LMArena Vision | 1172 | Not measured | Only Mistral Medium 3.1 measured |
| LMArena Text: English | 1438 | 1283 | Mistral Medium 3.1 |
| LMArena Text: non-English | 1408 | 1252 | Mistral Medium 3.1 |
| LMArena Text: Chinese | 1447 | 1273 | Mistral Medium 3.1 |
| LMArena Text: German | 1432 | 1234 | Mistral Medium 3.1 |
| LMArena Text: Japanese | 1379 | 1180 | Mistral Medium 3.1 |
| LMArena Text: Korean | 1380 | 1188 | Mistral Medium 3.1 |
| LMArena Text: French | 1460 | 1280 | Mistral Medium 3.1 |
| LMArena Text: Spanish | 1433 | 1255 | Mistral Medium 3.1 |
| LMArena Text: Russian | 1411 | 1264 | Mistral Medium 3.1 |
| LMArena Text: Polish | 1430 | Not measured | Only Mistral Medium 3.1 measured |
| LMArena Text: multi-turn | 1418 | 1272 | Mistral Medium 3.1 |
| LMArena Text: longer queries | 1406 | 1281 | Mistral Medium 3.1 |
| LMArena Text: writing and language | 1395 | 1247 | Mistral Medium 3.1 |
| LMArena Text: entertainment and media | 1390 | 1211 | Mistral Medium 3.1 |
| LMArena Vision: English | 1179 | Not measured | Only Mistral Medium 3.1 measured |
| LMArena Vision: Chinese | 1190 | Not measured | Only Mistral Medium 3.1 measured |
| LMArena Vision: reading text | 1175 | Not measured | Only Mistral Medium 3.1 measured |
| LMArena Vision: humor | 1168 | Not measured | Only Mistral Medium 3.1 measured |
| LMArena Vision: captioning | 1145 | Not measured | Only Mistral Medium 3.1 measured |
| LMArena Vision: recognizing things | 1147 | Not measured | Only Mistral Medium 3.1 measured |
| LMArena Vision: creative writing | 1200 | Not measured | Only Mistral Medium 3.1 measured |
| Neither model has a result on the other 14 boards here | |||
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | Mistral Medium 3.1 | Qwen2.5 72B Instruct | Leader |
|---|---|---|---|
| Neither model has a result on 5 boards here | |||
What each workload costs
Three fixed token mixes at each model's listed price, so per-token rates become a bill.
Chat turn
1K input + 500 output tokens
Qwen2.5 72B Instruct costs less
Repository review
50K input + 3K output tokens
Qwen2.5 72B Instruct costs less
Agent loop
200K input + 10K output tokens
Qwen2.5 72B Instruct costs less
Listed OpenRouter rates, without cache or batch discounts. A workload that does not fit a model's context window would need splitting.
Specification differences
The leading value in a row is marked; a row with a missing value names no leader. Our two rows are marked in red.
| Specification | Mistral Medium 3.1 | Qwen2.5 72B Instruct |
|---|---|---|
| Maker | Mistral AI | Qwen |
| Weights | No public weights listed | Open weights |
| Released | August 13, 2025 | September 19, 2024 |
| Context window | 131K | 33K |
| Input $/Mtok | $0.40 | $0.36 |
| Output $/Mtok | $2.00 | $0.40 |
| Coding index | Not benchmarked | Not benchmarked |
| Independent measurements behind it | 0 | 0 |
| Cost per solved task | Not computable | Not computable |
| Source | OpenRouter listing | OpenRouter listing |
Questions
Is Mistral Medium 3.1 better than Qwen2.5 72B Instruct for coding?
Neither model has a coding score yet, so the two cannot be ranked on coding. Pick Mistral Medium 3.1 for a larger context window (131K against 33K tokens) and work measured by MMLU Pro (Artificial Analysis run), where it leads.
Which is cheaper, Mistral Medium 3.1 or Qwen2.5 72B Instruct?
Mistral Medium 3.1 costs $0.40 per million input tokens and $2.00 per million output tokens; Qwen2.5 72B Instruct costs $0.36 and $0.40. Cost per solved task is known for neither of them.
Which has the bigger context window, Mistral Medium 3.1 or Qwen2.5 72B Instruct?
Mistral Medium 3.1, with 131K tokens against 33K.
How many benchmarks is the Mistral Medium 3.1 vs Qwen2.5 72B Instruct comparison based on?
39 benchmarks have a result for both models as of October 21, 2026. Mistral Medium 3.1's coding index rests on 0 independent benchmarks and Qwen2.5 72B Instruct's on 0. Where only one model has a result, the page says so instead of naming a winner.