Enter opens the highlighted result. Esc closes.

Mistral Medium 3 vs Qwen2.5 72B Instruct

Mistral Medium 3 leads Qwen2.5 72B Instruct on the Artificial Analysis Intelligence Index, 9.0 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.

Replace Mistral Medium 3 Replace Qwen2.5 72B Instruct
Mistral Medium 3Mistral AI
Coding
No data
Overall
9.0
Output $/M
$2.00
36
Qwen2.5 72B InstructQwen
Coding
No data
Overall
7.7
Output $/M
$0.40
3
vs boards won of 39 shared

Mistral Medium 3 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 for a larger context window (131K against 33K tokens) and work measured by Terminal-Bench Hard (Artificial Analysis run) or τ²-Bench Telecom, where it leads.

Model AMistral Medium 3Mistral AINot benchmarkedNo coding benchmarkNo cost per solved taskModel BQwen2.5 72B InstructQwenNot benchmarkedNo coding benchmarkNo cost per solved task
Shared results
39
Mistral Medium 3 only
18
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.

At a glance

Coding index
Mistral Medium 3 not benchmarked, Qwen2.5 72B Instruct not benchmarked.
Widest category
General, 16.1 points apart on average over 18 shared benchmarks.
Widest single gap
LMArena Text: German: Mistral Medium 3 1377 against Qwen2.5 72B Instruct 1234, 23.9 points apart on a 0 to 100 scale.
Output price
Mistral Medium 3 $2.00 against Qwen2.5 72B Instruct $0.40 per million output tokens, 5.0x apart.
Context window
Mistral Medium 3 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.

Boards won, where both have a result
CategoryMistral Medium 3BarsQwen2.5 72B InstructTies
Coding31
Intelligence20
Agentic work01
Reasoning50
Math40
Knowledge50
General171
All shared boards363
CodingLike-for-likeSix of these boards feed the Vibe Coding Index31
Coding benchmarks for Mistral Medium 3 and Qwen2.5 72B Instruct
BenchmarkMistral Medium 3Qwen2.5 72B InstructLeader
LiveCodeBench (Artificial Analysis run)40.0%27.6%Mistral Medium 3
Terminal-Bench Hard (Artificial Analysis run)3.8%4.5%Qwen2.5 72B Instruct
WeirdMLNot measured16.0%Only Qwen2.5 72B Instruct measured
LMArena Text: coding13851292Mistral Medium 3
LMArena Text: software and IT13851292Mistral Medium 3
Neither model has a result on the other 33 boards here
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index20
Intelligence benchmarks for Mistral Medium 3 and Qwen2.5 72B Instruct
BenchmarkMistral Medium 3Qwen2.5 72B InstructLeader
AA Intelligence Index9.07.7Mistral Medium 3
Epoch Capabilities Index134.0129.0Mistral Medium 3
Agentic workLike-for-likePublic boards, shown as published, kept out of the index01
Agentic work benchmarks for Mistral Medium 3 and Qwen2.5 72B Instruct
BenchmarkMistral Medium 3Qwen2.5 72B InstructLeader
τ²-Bench Telecom24.3%34.5%Qwen2.5 72B Instruct
METR time horizonsNot measured35.8%Only Qwen2.5 72B Instruct measured
BALROGNot measured16.2%Only Qwen2.5 72B Instruct measured
TheAgentCompanyNot measured5.7%Only Qwen2.5 72B Instruct measured
Neither model has a result on the other 35 boards here
ReasoningLike-for-likePublic boards, shown as published, kept out of the index50
Reasoning benchmarks for Mistral Medium 3 and Qwen2.5 72B Instruct
BenchmarkMistral Medium 3Qwen2.5 72B InstructLeader
GPQA Diamond (Artificial Analysis run)57.8%49.1%Mistral Medium 3
Humanity's Last Exam (Artificial Analysis run)4.1%3.6%Mistral Medium 3
GPQA Diamond (Epoch AI run)Not measured49.1%Only Qwen2.5 72B Instruct measured
BIG-Bench HardNot measured79.8%Only Qwen2.5 72B Instruct measured
LMArena Text: hard prompts13641271Mistral Medium 3
LMArena Text: expert13421245Mistral Medium 3
LMArena Text: hard prompts in English13741279Mistral Medium 3
LMArena Vision: diagrams1155Not measuredOnly Mistral Medium 3 measured
LMArena Vision: homework1165Not measuredOnly Mistral Medium 3 measured
Neither model has a result on the other 12 boards here
MathLike-for-likePublic boards, shown as published, kept out of the index40
Math benchmarks for Mistral Medium 3 and Qwen2.5 72B Instruct
BenchmarkMistral Medium 3Qwen2.5 72B InstructLeader
AA Math Index30.314.0Mistral Medium 3
AIME 2025 (Artificial Analysis run)30.3%14.0%Mistral Medium 3
OTIS Mock AIMENot measured8.1%Only Qwen2.5 72B Instruct measured
MATH level 5Not measured63.2%Only Qwen2.5 72B Instruct measured
LMArena Text: math13531283Mistral Medium 3
LMArena Text: mathematical work13541290Mistral Medium 3
Neither model has a result on the other 7 boards here
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index50
Knowledge benchmarks for Mistral Medium 3 and Qwen2.5 72B Instruct
BenchmarkMistral Medium 3Qwen2.5 72B InstructLeader
MMLU Pro (Artificial Analysis run)76.0%72.0%Mistral Medium 3
MMLUNot measured85.3%Only Qwen2.5 72B Instruct measured
AI2 Reasoning ChallengeNot measured94.5%Only Qwen2.5 72B Instruct measured
LMArena Text: science13811275Mistral Medium 3
LMArena Text: medicine and healthcare13811258Mistral Medium 3
LMArena Text: legal and government13841288Mistral Medium 3
LMArena Text: business and finance13521253Mistral Medium 3
Neither model has a result on the other 10 boards here
Web designDirectional onlyPublic boards, shown as published, kept out of the index6 boards, none shared
Web design benchmarks for Mistral Medium 3 and Qwen2.5 72B Instruct
BenchmarkMistral Medium 3Qwen2.5 72B InstructLeader
Design Arena: websites1079Not measuredOnly Mistral Medium 3 measured
Design Arena: UI components1024Not measuredOnly Mistral Medium 3 measured
Design Arena: game dev1018Not measuredOnly Mistral Medium 3 measured
Design Arena: data viz1034Not measuredOnly Mistral Medium 3 measured
Design Arena: 3D1084Not measuredOnly Mistral Medium 3 measured
Design Arena: all code categories1069Not measuredOnly Mistral Medium 3 measured
Neither model has a result on the other 3 boards here
GeneralLike-for-likePublic boards, shown as published, kept out of the index171
General benchmarks for Mistral Medium 3 and Qwen2.5 72B Instruct
BenchmarkMistral Medium 3Qwen2.5 72B InstructLeader
IFBench39.3%36.9%Mistral Medium 3
Long context reasoning (AA-LCR)31.3%Not measuredOnly Mistral Medium 3 measured
DTBench62.3%62.9%Qwen2.5 72B Instruct
LMCANot measured13.4%Only Qwen2.5 72B Instruct measured
HellaSwagNot measured84.8%Only Qwen2.5 72B Instruct measured
WinoGrandeNot measured82.3%Only Qwen2.5 72B Instruct measured
PIQANot measured82.6%Only Qwen2.5 72B Instruct measured
LMArena Text13701269Mistral Medium 3
LMArena Text: instruction following13381254Mistral Medium 3
LMArena Text: creative writing13431221Mistral Medium 3
LMArena Vision1157Not measuredOnly Mistral Medium 3 measured
LMArena Text: English13851283Mistral Medium 3
LMArena Text: non-English13531252Mistral Medium 3
LMArena Text: Chinese13741273Mistral Medium 3
LMArena Text: German13771234Mistral Medium 3
LMArena Text: Japanese13161180Mistral Medium 3
LMArena Text: Korean13041188Mistral Medium 3
LMArena Text: French13681280Mistral Medium 3
LMArena Text: Spanish13641255Mistral Medium 3
LMArena Text: Russian13591264Mistral Medium 3
LMArena Text: Polish1371Not measuredOnly Mistral Medium 3 measured
LMArena Text: multi-turn13841272Mistral Medium 3
LMArena Text: longer queries13601281Mistral Medium 3
LMArena Text: writing and language13521247Mistral Medium 3
LMArena Text: entertainment and media13351211Mistral Medium 3
LMArena Vision: English1172Not measuredOnly Mistral Medium 3 measured
LMArena Vision: Chinese1161Not measuredOnly Mistral Medium 3 measured
LMArena Vision: reading text1154Not measuredOnly Mistral Medium 3 measured
LMArena Vision: humor1145Not measuredOnly Mistral Medium 3 measured
LMArena Vision: captioning1094Not measuredOnly Mistral Medium 3 measured
LMArena Vision: recognizing things1141Not measuredOnly Mistral Medium 3 measured
LMArena Vision: creative writing1146Not measuredOnly Mistral Medium 3 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
Image generation benchmarks for Mistral Medium 3 and Qwen2.5 72B Instruct
BenchmarkMistral Medium 3Qwen2.5 72B InstructLeader
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

Mistral Medium 3$0.0014Fits in one request
Qwen2.5 72B Instruct$0.0006Fits in one request

Qwen2.5 72B Instruct costs less

Repository review

50K input + 3K output tokens

Mistral Medium 3$0.026Fits in one request
Qwen2.5 72B Instruct$0.019Exceeds the context window

Qwen2.5 72B Instruct costs less

Agent loop

200K input + 10K output tokens

Mistral Medium 3$0.100Exceeds the context window
Qwen2.5 72B Instruct$0.076Exceeds the context window

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.

Specifications of Mistral Medium 3 and Qwen2.5 72B Instruct
SpecificationMistral Medium 3Qwen2.5 72B Instruct
MakerMistral AIQwen
WeightsNo public weights listedOpen weights
ReleasedMay 7, 2025September 19, 2024
Context window131K33K
Input $/Mtok$0.40$0.36
Output $/Mtok$2.00$0.40
Coding indexNot benchmarkedNot benchmarked
Independent measurements behind it00
Cost per solved taskNot computableNot computable
SourceOpenRouter listingOpenRouter listing

Questions

Is Mistral Medium 3 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 for a larger context window (131K against 33K tokens) and work measured by Terminal-Bench Hard (Artificial Analysis run) or τ²-Bench Telecom, where it leads.

Which is cheaper, Mistral Medium 3 or Qwen2.5 72B Instruct?

Mistral Medium 3 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 or Qwen2.5 72B Instruct?

Mistral Medium 3, with 131K tokens against 33K.

How many benchmarks is the Mistral Medium 3 vs Qwen2.5 72B Instruct comparison based on?

39 benchmarks have a result for both models as of October 21, 2026. Mistral Medium 3'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.

Compare next