Enter opens the highlighted result. Esc closes.

Phi 4 vs Mistral Medium 3

Updated October 21, 2026. 39 benchmarks have a result for both. Coding is ranked on the Vibe Coding Index; every other board is shown as published.

Replace Phi 4 Replace Mistral Medium 3
Phi 4Microsoft
Coding
No data
Overall
5.9
Output $/M
$0.14
0
Mistral Medium 3Mistral AI
Coding
No data
Overall
9.0
Output $/M
$2.00
38
vs boards won of 39 shared, 1 tied

Decision reading

Neither model has a coding score yet, so the two cannot be ranked on coding.

Phi 4 has the edge where output tokens cost 14.3x less, and where what matters is LiveCodeBench (Artificial Analysis run) or LMArena Text: coding, where it leads.

Model APhi 4MicrosoftNot benchmarkedNo coding benchmarkNo cost per solved taskModel BMistral Medium 3Mistral AINot benchmarkedNo coding benchmarkNo cost per solved task
Shared results
39
Phi 4 only
6
Mistral Medium 3 only
18
Like-for-like categories
7 of 9

Shared results can support a head-to-head reading. Results only one model has describe coverage, not superiority.

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
Phi 4 not benchmarked, Mistral Medium 3 not benchmarked.
Widest category
General, 25.4 points apart on average over 18 shared benchmarks.
Widest single gap
Long context reasoning (AA-LCR): Phi 4 0.0% against Mistral Medium 3 31.3%, 31.3 points apart on a 0 to 100 scale.
Output price
Phi 4 $0.14 against Mistral Medium 3 $2.00 per million output tokens, 14.3x apart.
Context window
Phi 4 takes 16K tokens, Mistral Medium 3 131K.

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
CategoryPhi 4BarsMistral Medium 3Ties
Coding031 tie
Intelligence02
Agentic work01
Reasoning05
Math04
Knowledge05
General018
All shared boards0381 tie
CodingLike-for-likeSix of these boards feed the Vibe Coding Index03
Coding benchmarks for Phi 4 and Mistral Medium 3
BenchmarkPhi 4Mistral Medium 3Leader
LiveCodeBench (Artificial Analysis run)23.1%40.0%Mistral Medium 3
Terminal-Bench Hard (Artificial Analysis run)3.8%3.8%Tie
LMArena Text: coding12311385Mistral Medium 3
LMArena Text: software and IT12281385Mistral Medium 3
Neither model has a result on the other 34 boards here
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index02
Intelligence benchmarks for Phi 4 and Mistral Medium 3
BenchmarkPhi 4Mistral Medium 3Leader
AA Intelligence Index5.99.0Mistral Medium 3
Epoch Capabilities Index130.5134.0Mistral Medium 3
Agentic workLike-for-likePublic boards, shown as published, kept out of the index01
Agentic work benchmarks for Phi 4 and Mistral Medium 3
BenchmarkPhi 4Mistral Medium 3Leader
τ²-Bench Telecom0.0%24.3%Mistral Medium 3
BALROG11.6%Not measuredOnly Phi 4 measured
Neither model has a result on the other 37 boards here
ReasoningLike-for-likePublic boards, shown as published, kept out of the index05
Reasoning benchmarks for Phi 4 and Mistral Medium 3
BenchmarkPhi 4Mistral Medium 3Leader
GPQA Diamond (Artificial Analysis run)57.5%57.8%Mistral Medium 3
Humanity's Last Exam (Artificial Analysis run)3.8%4.1%Mistral Medium 3
GPQA Diamond (Epoch AI run)56.1%Not measuredOnly Phi 4 measured
LMArena Text: hard prompts12201364Mistral Medium 3
LMArena Text: expert12031342Mistral Medium 3
LMArena Text: hard prompts in English12301374Mistral Medium 3
LMArena Vision: diagramsNot measured1155Only Mistral Medium 3 measured
LMArena Vision: homeworkNot measured1165Only Mistral Medium 3 measured
Neither model has a result on the other 13 boards here
MathLike-for-likePublic boards, shown as published, kept out of the index04
Math benchmarks for Phi 4 and Mistral Medium 3
BenchmarkPhi 4Mistral Medium 3Leader
AA Math Index18.030.3Mistral Medium 3
AIME 2025 (Artificial Analysis run)18.0%30.3%Mistral Medium 3
OTIS Mock AIME13.8%Not measuredOnly Phi 4 measured
MATH level 564.9%Not measuredOnly Phi 4 measured
LMArena Text: math12461353Mistral Medium 3
LMArena Text: mathematical work12541354Mistral Medium 3
Neither model has a result on the other 7 boards here
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index05
Knowledge benchmarks for Phi 4 and Mistral Medium 3
BenchmarkPhi 4Mistral Medium 3Leader
MMLU Pro (Artificial Analysis run)71.4%76.0%Mistral Medium 3
MMLU84.8%Not measuredOnly Phi 4 measured
LMArena Text: science12061381Mistral Medium 3
LMArena Text: medicine and healthcare12021381Mistral Medium 3
LMArena Text: legal and government12411384Mistral Medium 3
LMArena Text: business and finance11991352Mistral Medium 3
Neither model has a result on the other 11 boards here
Web designDirectional onlyPublic boards, shown as published, kept out of the index6 boards, none shared
Web design benchmarks for Phi 4 and Mistral Medium 3
BenchmarkPhi 4Mistral Medium 3Leader
Design Arena: websitesNot measured1079Only Mistral Medium 3 measured
Design Arena: UI componentsNot measured1024Only Mistral Medium 3 measured
Design Arena: game devNot measured1018Only Mistral Medium 3 measured
Design Arena: data vizNot measured1034Only Mistral Medium 3 measured
Design Arena: 3DNot measured1083Only Mistral Medium 3 measured
Design Arena: all code categoriesNot measured1069Only 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 index018
General benchmarks for Phi 4 and Mistral Medium 3
BenchmarkPhi 4Mistral Medium 3Leader
IFBench23.5%39.3%Mistral Medium 3
Long context reasoning (AA-LCR)0.0%31.3%Mistral Medium 3
DTBenchNot measured62.3%Only Mistral Medium 3 measured
Creative writing (Lech Mazur)62.6%Not measuredOnly Phi 4 measured
LMArena Text12171370Mistral Medium 3
LMArena Text: instruction following12011338Mistral Medium 3
LMArena Text: creative writing11811343Mistral Medium 3
LMArena VisionNot measured1157Only Mistral Medium 3 measured
LMArena Text: English12311385Mistral Medium 3
LMArena Text: non-English11971353Mistral Medium 3
LMArena Text: Chinese12121374Mistral Medium 3
LMArena Text: German12221377Mistral Medium 3
LMArena Text: Japanese11581316Mistral Medium 3
LMArena Text: Korean11511304Mistral Medium 3
LMArena Text: French12241368Mistral Medium 3
LMArena Text: Spanish12341364Mistral Medium 3
LMArena Text: Russian12091359Mistral Medium 3
LMArena Text: PolishNot measured1371Only Mistral Medium 3 measured
LMArena Text: multi-turn12061384Mistral Medium 3
LMArena Text: longer queries12171360Mistral Medium 3
LMArena Text: writing and language11981352Mistral Medium 3
LMArena Text: entertainment and media11621335Mistral Medium 3
LMArena Vision: EnglishNot measured1172Only Mistral Medium 3 measured
LMArena Vision: ChineseNot measured1161Only Mistral Medium 3 measured
LMArena Vision: reading textNot measured1154Only Mistral Medium 3 measured
LMArena Vision: humorNot measured1145Only Mistral Medium 3 measured
LMArena Vision: captioningNot measured1094Only Mistral Medium 3 measured
LMArena Vision: recognizing thingsNot measured1141Only Mistral Medium 3 measured
LMArena Vision: creative writingNot measured1146Only Mistral Medium 3 measured
Neither model has a result on the other 17 boards here
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
Image generation benchmarks for Phi 4 and Mistral Medium 3
BenchmarkPhi 4Mistral Medium 3Leader
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

Phi 4$0.0001Fits in one request
Mistral Medium 3$0.0014Fits in one request

Phi 4 costs less

Repository review

50K input + 3K output tokens

Phi 4$0.0039Exceeds the context window
Mistral Medium 3$0.026Fits in one request

Phi 4 costs less

Agent loop

200K input + 10K output tokens

Phi 4$0.015Exceeds the context window
Mistral Medium 3$0.100Exceeds the context window

Phi 4 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 Phi 4 and Mistral Medium 3
SpecificationMicrosoft Phi 4Mistral Medium 3
MakerMicrosoftMistral AI
WeightsOpen weightsNo public weights listed
ReleasedJanuary 10, 2025May 7, 2025
Context window16K131K
Input $/Mtok$0.07$0.40
Output $/Mtok$0.14$2.00
Coding indexNot benchmarkedNot benchmarked
Independent measurements behind it00
Cost per solved taskNot computableNot computable
SourceOpenRouter listingOpenRouter listing

Questions

Is Phi 4 better than Mistral Medium 3 for coding?

Neither model has a coding score yet, so the two cannot be ranked on coding. Phi 4 has the edge where output tokens cost 14.3x less, and where what matters is LiveCodeBench (Artificial Analysis run) or LMArena Text: coding, where it leads.

Which is cheaper, Phi 4 or Mistral Medium 3?

Phi 4 costs $0.07 per million input tokens and $0.14 per million output tokens; Mistral Medium 3 costs $0.40 and $2.00. Cost per solved task is known for neither of them.

Which has the bigger context window, Phi 4 or Mistral Medium 3?

Mistral Medium 3, with 131K tokens against 16K.

How many benchmarks is the Phi 4 vs Mistral Medium 3 comparison based on?

39 benchmarks have a result for both models as of October 21, 2026. Phi 4's coding index rests on 0 independent benchmarks and Mistral Medium 3's on 0. Where only one model has a result, the page says so instead of naming a winner.

Compare next