Enter opens the highlighted result. Esc closes.

Phi 4 vs MiniMax M2

Updated October 21, 2026. 36 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 MiniMax M2
Phi 4Microsoft
Coding
No data
Overall
5.9
Output $/M
$0.14
0
MiniMax M2MiniMax
Coding
No data
Overall
18.6
Output $/M
$1.02
36
vs boards won of 36 shared

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 7.3x less, and where what matters is LiveCodeBench (Artificial Analysis run) or Terminal-Bench Hard (Artificial Analysis run), where it leads.

Model APhi 4MicrosoftNot benchmarkedNo coding benchmarkNo cost per solved taskModel BMiniMax M2MiniMaxNot benchmarked1 independent benchmark$0.702 per solved task
Shared results
36
Phi 4 only
9
MiniMax M2 only
13
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, MiniMax M2 not benchmarked.
Widest category
Agentic work, 86.8 points apart on average over 1 shared benchmark.
Widest single gap
τ²-Bench Telecom: Phi 4 0.0% against MiniMax M2 86.8%, 86.8 points apart on a 0 to 100 scale.
Output price
Phi 4 $0.14 against MiniMax M2 $1.02 per million output tokens, 7.3x apart.
Context window
Phi 4 takes 16K tokens, MiniMax M2 205K.

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 4BarsMiniMax M2Ties
Coding04
Intelligence01
Agentic work01
Reasoning05
Math04
Knowledge05
General016
All shared boards036
CodingLike-for-likeSix of these boards feed the Vibe Coding Index04
Coding benchmarks for Phi 4 and MiniMax M2
BenchmarkPhi 4MiniMax M2Leader
SWE-bench VerifiedNot measured61.0%Only MiniMax M2 measured
LMArena WebDevNot measured1297Only MiniMax M2 measured
LiveCodeBench (Artificial Analysis run)23.1%82.6%MiniMax M2
Terminal-Bench Hard (Artificial Analysis run)3.8%25.8%MiniMax M2
Terminal-BenchNot measured30.0%Only MiniMax M2 measured
LMArena Text: coding12311371MiniMax M2
LMArena Text: software and IT12281373MiniMax M2
LMArena WebDev: HTMLNot measured1306Only MiniMax M2 measured
Neither model has a result on the other 30 boards here
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index01
Intelligence benchmarks for Phi 4 and MiniMax M2
BenchmarkPhi 4MiniMax M2Leader
AA Intelligence Index5.918.6MiniMax M2
Epoch Capabilities Index130.5Not measuredOnly Phi 4 measured
Agentic workLike-for-likePublic boards, shown as published, kept out of the index01
Agentic work benchmarks for Phi 4 and MiniMax M2
BenchmarkPhi 4MiniMax M2Leader
τ²-Bench Telecom0.0%86.8%MiniMax M2
BALROG11.6%Not measuredOnly Phi 4 measured
Vending-Bench 2Not measured$161Only MiniMax M2 measured
Neither model has a result on the other 36 boards here
ReasoningLike-for-likePublic boards, shown as published, kept out of the index05
Reasoning benchmarks for Phi 4 and MiniMax M2
BenchmarkPhi 4MiniMax M2Leader
GPQA Diamond (Artificial Analysis run)57.5%77.7%MiniMax M2
Humanity's Last Exam (Artificial Analysis run)3.8%13.7%MiniMax M2
GPQA Diamond (Epoch AI run)56.1%Not measuredOnly Phi 4 measured
LMArena Text: hard prompts12201357MiniMax M2
LMArena Text: expert12031338MiniMax M2
LMArena Text: hard prompts in English12301376MiniMax M2
Neither model has a result on the other 15 boards here
MathLike-for-likePublic boards, shown as published, kept out of the index04
Math benchmarks for Phi 4 and MiniMax M2
BenchmarkPhi 4MiniMax M2Leader
AA Math Index18.078.3MiniMax M2
AIME 2025 (Artificial Analysis run)18.0%78.3%MiniMax M2
OTIS Mock AIME13.8%Not measuredOnly Phi 4 measured
MATH level 564.9%Not measuredOnly Phi 4 measured
LMArena Text: math12461351MiniMax M2
LMArena Text: mathematical work12541368MiniMax M2
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 MiniMax M2
BenchmarkPhi 4MiniMax M2Leader
MMLU Pro (Artificial Analysis run)71.4%82.0%MiniMax M2
MMLU84.8%Not measuredOnly Phi 4 measured
LMArena Text: science12061348MiniMax M2
LMArena Text: medicine and healthcare12021367MiniMax M2
LMArena Text: legal and government12411349MiniMax M2
LMArena Text: business and finance11991336MiniMax M2
Neither model has a result on the other 11 boards here
Web designDirectional onlyPublic boards, shown as published, kept out of the index7 boards, none shared
Web design benchmarks for Phi 4 and MiniMax M2
BenchmarkPhi 4MiniMax M2Leader
Design Arena: websitesNot measured1143Only MiniMax M2 measured
Design Arena: UI componentsNot measured1136Only MiniMax M2 measured
Design Arena: game devNot measured1118Only MiniMax M2 measured
Design Arena: data vizNot measured1140Only MiniMax M2 measured
Design Arena: 3DNot measured1098Only MiniMax M2 measured
Design Arena: all code categoriesNot measured1135Only MiniMax M2 measured
Design Arena: SVGNot measured1100Only MiniMax M2 measured
Neither model has a result on the other 2 boards here
GeneralLike-for-likePublic boards, shown as published, kept out of the index016
General benchmarks for Phi 4 and MiniMax M2
BenchmarkPhi 4MiniMax M2Leader
IFBench23.5%72.3%MiniMax M2
Long context reasoning (AA-LCR)0.0%64.3%MiniMax M2
Creative writing (Lech Mazur)62.6%Not measuredOnly Phi 4 measured
LMArena Text12171340MiniMax M2
LMArena Text: instruction following12011329MiniMax M2
LMArena Text: creative writing11811286MiniMax M2
LMArena Text: English12311364MiniMax M2
LMArena Text: non-English11971313MiniMax M2
LMArena Text: Chinese12121366MiniMax M2
LMArena Text: German12221357MiniMax M2
LMArena Text: Japanese1158Not measuredOnly Phi 4 measured
LMArena Text: Korean1151Not measuredOnly Phi 4 measured
LMArena Text: French12241336MiniMax M2
LMArena Text: Spanish12341326MiniMax M2
LMArena Text: Russian12091330MiniMax M2
LMArena Text: PolishNot measured1312Only MiniMax M2 measured
LMArena Text: multi-turn12061361MiniMax M2
LMArena Text: longer queries12171331MiniMax M2
LMArena Text: writing and language11981295MiniMax M2
LMArena Text: entertainment and media11621298MiniMax M2
Neither model has a result on the other 26 boards here
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
Image generation benchmarks for Phi 4 and MiniMax M2
BenchmarkPhi 4MiniMax M2Leader
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
MiniMax M2$0.0008Fits in one request

Phi 4 costs less

Repository review

50K input + 3K output tokens

Phi 4$0.0039Exceeds the context window
MiniMax M2$0.016Fits in one request

Phi 4 costs less

Agent loop

200K input + 10K output tokens

Phi 4$0.015Exceeds the context window
MiniMax M2$0.061Exceeds 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 MiniMax M2
SpecificationMicrosoft Phi 4MiniMax M2
MakerMicrosoftMiniMax
WeightsOpen weightsOpen weights
ReleasedJanuary 10, 2025October 23, 2025
Context window16K205K
Input $/Mtok$0.07$0.255
Output $/Mtok$0.14$1.02
Coding indexNot benchmarkedNot benchmarked
Independent measurements behind it01
Cost per solved taskNot computable$0.702
SourceOpenRouter listingOpenRouter listing

Questions

Is Phi 4 better than MiniMax M2 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 7.3x less, and where what matters is LiveCodeBench (Artificial Analysis run) or Terminal-Bench Hard (Artificial Analysis run), where it leads.

Which is cheaper, Phi 4 or MiniMax M2?

Phi 4 costs $0.07 per million input tokens and $0.14 per million output tokens; MiniMax M2 costs $0.255 and $1.02. Cost per solved task is known for MiniMax M2 only.

Which has the bigger context window, Phi 4 or MiniMax M2?

MiniMax M2, with 205K tokens against 16K.

How many benchmarks is the Phi 4 vs MiniMax M2 comparison based on?

36 benchmarks have a result for both models as of October 21, 2026. Phi 4's coding index rests on 0 independent benchmarks and MiniMax M2's on 1. Where only one model has a result, the page says so instead of naming a winner.

Compare next