Enter opens the highlighted result. Esc closes.

Grok 4.20 vs MiniMax M2.7

MiniMax M2.7 leads Grok 4.20 on coding, 36.2 to 33.3 on the Vibe Coding Index, and MiniMax M2.7 costs 2.1x less per output token ($1.20 vs $2.50 per million). Updated October 22, 2026; 36 benchmarks have a result for both.

Replace Grok 4.20 Replace MiniMax M2.7
Grok 4.20xAI
Coding
33.3
Overall
41.0
Output $/M
$2.50
19
MiniMax M2.7MiniMax
Coding
36.2
Overall
44.5
Output $/M
$1.20
17
vs boards won of 36 shared

Grok 4.20 or MiniMax M2.7: which to pick

Pick MiniMax M2.7 for coding: it scores 36.2 on the Vibe Coding Index against 33.3 for Grok 4.20.

Pick Grok 4.20 for a larger context window (2M against 205K tokens) and work measured by LiveCodeBench (Vals AI) or Terminal-Bench, where it leads.

Model AGrok 4.20xAI33.3/1002 independent benchmarks$0.72 per solved taskModel BMiniMax M2.7MiniMax36.2/1003 independent benchmarks$0.64 per solved task
Shared results
36
Grok 4.20 only
19
MiniMax M2.7 only
38
Like-for-like categories
8 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
Grok 4.20 33.3 from 2 benchmarks, MiniMax M2.7 36.2 from 3 benchmarks.
Widest category
General, 10.4 points apart on average over 3 shared benchmarks.
Widest single gap
Excel Modeling Benchmark (Vals AI): Grok 4.20 11.7% against MiniMax M2.7 28.4%, 16.7 points apart on a 0 to 100 scale.
Output price
Grok 4.20 $2.50 against MiniMax M2.7 $1.20 per million output tokens, 2.1x apart.
Context window
Grok 4.20 takes 2M tokens, MiniMax M2.7 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
CategoryGrok 4.20BarsMiniMax M2.7Ties
Coding35
Intelligence20
Agentic work41
Reasoning30
Math10
Knowledge33
Web design26
General12
All shared boards1917
CodingLike-for-likeSix of these boards feed the Vibe Coding Index35
Coding benchmarks for Grok 4.20 and MiniMax M2.7
BenchmarkGrok 4.20MiniMax M2.7Leader
LMArena WebDevNot measured1399Only MiniMax M2.7 measured
AA Coding IndexNot measured52.6Only MiniMax M2.7 measured
LiveCodeBench (Vals AI)84.3%79.9%Grok 4.20
SWE-bench Verified (Vals AI)72.2%73.8%MiniMax M2.7
Vibe Code Bench (Vals AI)4.1%11.9%MiniMax M2.7
Code migration (Vals AI)0.3%8.7%MiniMax M2.7
ProgramBench (Vals AI)Not measured0.0%Only MiniMax M2.7 measured
Terminal-Bench 2.1 (Vals AI)44.2%48.7%MiniMax M2.7
SciCodeNot measured50.1%Only MiniMax M2.7 measured
Terminal-Bench Hard (Artificial Analysis run)37.9%39.4%MiniMax M2.7
Terminal-Bench57.3%45.1%Grok 4.20
WeirdML52.3%37.0%Grok 4.20
ALE-BenchNot measured599.3Only MiniMax M2.7 measured
LMArena Text: codingNot measured1454Only MiniMax M2.7 measured
LMArena Text: software and ITNot measured1446Only MiniMax M2.7 measured
LMArena WebDev: ReactNot measured1388Only MiniMax M2.7 measured
LMArena WebDev: HTMLNot measured1401Only MiniMax M2.7 measured
Neither model has a result on the other 26 boards here
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index20
Intelligence benchmarks for Grok 4.20 and MiniMax M2.7
BenchmarkGrok 4.20MiniMax M2.7Leader
AA Intelligence Index25.722.8Grok 4.20
Epoch Capabilities Index152.1145.8Grok 4.20
Agentic workLike-for-likePublic boards, shown as published, kept out of the index41
Agentic work benchmarks for Grok 4.20 and MiniMax M2.7
BenchmarkGrok 4.20MiniMax M2.7Leader
AA Agentic IndexNot measured15.3Only MiniMax M2.7 measured
Design Arena: agents building web apps1144Not measuredOnly Grok 4.20 measured
Design Arena: agents building full-stack apps1054Not measuredOnly Grok 4.20 measured
Design Arena: agents building mobile apps1090Not measuredOnly Grok 4.20 measured
Design Arena: agents building Android apps1045Not measuredOnly Grok 4.20 measured
Design Arena: agents building Godot games1089Not measuredOnly Grok 4.20 measured
Design Arena: HTML slides1180Not measuredOnly Grok 4.20 measured
Finance Agent (Vals AI)28.5%27.9%Grok 4.20
Tax Agent Bench (Vals AI)7.0%6.9%Grok 4.20
Legal Research Bench (Vals AI)13.9%10.6%Grok 4.20
Excel Modeling Benchmark (Vals AI)11.7%28.4%MiniMax M2.7
τ²-Bench Telecom93.0%84.8%Grok 4.20
Vending-Bench 2$4,663Not measuredOnly Grok 4.20 measured
Neither model has a result on the other 29 boards here
ReasoningLike-for-likePublic boards, shown as published, kept out of the index30
Reasoning benchmarks for Grok 4.20 and MiniMax M2.7
BenchmarkGrok 4.20MiniMax M2.7Leader
GPQA Diamond88.6%86.6%Grok 4.20
MMMU Pro (Vals AI)83.5%Not measuredOnly Grok 4.20 measured
GPQA Diamond (Artificial Analysis run)91.1%87.4%Grok 4.20
Humanity's Last Exam (Artificial Analysis run)34.5%29.6%Grok 4.20
ARC-AGI89.5%Not measuredOnly Grok 4.20 measured
ARC-AGI-265.1%Not measuredOnly Grok 4.20 measured
LMArena Text: hard promptsNot measured1422Only MiniMax M2.7 measured
LMArena Text: expertNot measured1443Only MiniMax M2.7 measured
LMArena Text: hard prompts in EnglishNot measured1441Only MiniMax M2.7 measured
Kagi LLM Benchmark75.0%Not measuredOnly Grok 4.20 measured
Neither model has a result on the other 13 boards here
MathLike-for-likePublic boards, shown as published, kept out of the index10
Math benchmarks for Grok 4.20 and MiniMax M2.7
BenchmarkGrok 4.20MiniMax M2.7Leader
AIME96.5%91.0%Grok 4.20
ProofBenchNot measured3.0%Only MiniMax M2.7 measured
LMArena Text: mathNot measured1421Only MiniMax M2.7 measured
LMArena Text: mathematical workNot measured1437Only MiniMax M2.7 measured
Neither model has a result on the other 10 boards here
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index33
Knowledge benchmarks for Grok 4.20 and MiniMax M2.7
BenchmarkGrok 4.20MiniMax M2.7Leader
MMLU Pro86.3%80.4%Grok 4.20
LegalBench (Vals AI)77.7%84.0%MiniMax M2.7
CaseLaw (Vals AI)54.4%60.9%MiniMax M2.7
CorpFin (Vals AI)63.7%61.2%Grok 4.20
TaxEval (Vals AI)74.1%66.6%Grok 4.20
MortgageTax (Vals AI)45.4%Not measuredOnly Grok 4.20 measured
MedQA (Vals AI)94.5%Not measuredOnly Grok 4.20 measured
MedCode (Vals AI)32.2%34.4%MiniMax M2.7
LMArena Text: scienceNot measured1416Only MiniMax M2.7 measured
LMArena Text: medicine and healthcareNot measured1409Only MiniMax M2.7 measured
LMArena Text: legal and governmentNot measured1403Only MiniMax M2.7 measured
LMArena Text: business and financeNot measured1407Only MiniMax M2.7 measured
Neither model has a result on the other 8 boards here
Web designLike-for-likePublic boards, shown as published, kept out of the index26
Web design benchmarks for Grok 4.20 and MiniMax M2.7
BenchmarkGrok 4.20MiniMax M2.7Leader
Design Arena: websites12291242MiniMax M2.7
Design Arena: UI components11971207MiniMax M2.7
Design Arena: game dev11931209MiniMax M2.7
Design Arena: data viz12171233MiniMax M2.7
Design Arena: 3D11931194MiniMax M2.7
Design Arena: all code categories12211233MiniMax M2.7
Design Arena: SVG11651137Grok 4.20
Design Arena: ASCII art11751136Grok 4.20
Neither model has a result on the other 1 board here
GeneralLike-for-likePublic boards, shown as published, kept out of the index12
General benchmarks for Grok 4.20 and MiniMax M2.7
BenchmarkGrok 4.20MiniMax M2.7Leader
SAGE (Vals AI)38.2%Not measuredOnly Grok 4.20 measured
MedScribe (Vals AI)63.4%79.9%MiniMax M2.7
Vals Multimodal Index39.1%Not measuredOnly Grok 4.20 measured
IFBench81.2%75.7%Grok 4.20
Long context reasoning (AA-LCR)69.0%78.3%MiniMax M2.7
DTBench90.1%Not measuredOnly Grok 4.20 measured
LMCA38.7%Not measuredOnly Grok 4.20 measured
CL-bench22.2%Not measuredOnly Grok 4.20 measured
ExploitBenchNot measured13.3%Only MiniMax M2.7 measured
CL-bench Life11.9%Not measuredOnly Grok 4.20 measured
LMArena TextNot measured1405Only MiniMax M2.7 measured
LMArena Text: instruction followingNot measured1405Only MiniMax M2.7 measured
LMArena Text: creative writingNot measured1354Only MiniMax M2.7 measured
LMArena Text: EnglishNot measured1426Only MiniMax M2.7 measured
LMArena Text: non-EnglishNot measured1382Only MiniMax M2.7 measured
LMArena Text: ChineseNot measured1442Only MiniMax M2.7 measured
LMArena Text: GermanNot measured1398Only MiniMax M2.7 measured
LMArena Text: JapaneseNot measured1260Only MiniMax M2.7 measured
LMArena Text: KoreanNot measured1314Only MiniMax M2.7 measured
LMArena Text: FrenchNot measured1424Only MiniMax M2.7 measured
LMArena Text: SpanishNot measured1402Only MiniMax M2.7 measured
LMArena Text: RussianNot measured1383Only MiniMax M2.7 measured
LMArena Text: PolishNot measured1387Only MiniMax M2.7 measured
LMArena Text: multi-turnNot measured1412Only MiniMax M2.7 measured
LMArena Text: longer queriesNot measured1418Only MiniMax M2.7 measured
LMArena Text: writing and languageNot measured1376Only MiniMax M2.7 measured
LMArena Text: entertainment and mediaNot measured1360Only MiniMax M2.7 measured
Neither model has a result on the other 21 boards here
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
Image generation benchmarks for Grok 4.20 and MiniMax M2.7
BenchmarkGrok 4.20MiniMax M2.7Leader
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

Grok 4.20$0.0025Fits in one request
MiniMax M2.7$0.0009Fits in one request

MiniMax M2.7 costs less

Repository review

50K input + 3K output tokens

Grok 4.20$0.070Fits in one request
MiniMax M2.7$0.019Fits in one request

MiniMax M2.7 costs less

Agent loop

200K input + 10K output tokens

Grok 4.20$0.275Fits in one request
MiniMax M2.7$0.072Exceeds the context window

MiniMax M2.7 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 Grok 4.20 and MiniMax M2.7
SpecificationSpaceXAI Grok 4.20MiniMax M2.7
MakerxAIMiniMax
WeightsNo public weights listedOpen weights
ReleasedMarch 31, 2026March 18, 2026
Context window2M205K
Input $/Mtok$1.25$0.30
Output $/Mtok$2.50$1.20
Coding index33.336.2
Independent measurements behind it23
Cost per solved task$0.72$0.64
SourceOpenRouter listingOpenRouter listing

Questions

Is Grok 4.20 better than MiniMax M2.7 for coding?

Pick MiniMax M2.7 for coding: it scores 36.2 on the Vibe Coding Index against 33.3 for Grok 4.20. Pick Grok 4.20 for a larger context window (2M against 205K tokens) and work measured by LiveCodeBench (Vals AI) or Terminal-Bench, where it leads.

Which is cheaper, Grok 4.20 or MiniMax M2.7?

Grok 4.20 costs $1.25 per million input tokens and $2.50 per million output tokens; MiniMax M2.7 costs $0.30 and $1.20. Per solved SWE-bench Verified task, Grok 4.20 costs $0.72 and MiniMax M2.7 $0.64.

Which has the bigger context window, Grok 4.20 or MiniMax M2.7?

Grok 4.20, with 2M tokens against 205K.

How many benchmarks is the Grok 4.20 vs MiniMax M2.7 comparison based on?

36 benchmarks have a result for both models as of October 22, 2026. Grok 4.20's coding index rests on 2 independent benchmarks and MiniMax M2.7's on 3. Where only one model has a result, the page says so instead of naming a winner.

Compare next

Embed this comparison on your site

Free under CC BY 4.0, while the source link stays visible. Preview the frame · Terms and other embeds

<iframe src="https://llmbenchmark.ai/embed/compare/grok-4-20-vs-minimax-minimax-m2-7" title="Grok 4.20 vs MiniMax M2.7, LLMBenchmark.ai" width="100%" height="360" style="border:0;max-width:100%" loading="lazy"></iframe>
<p style="font:13px/1.4 sans-serif;margin:6px 0 0">Source: <a href="https://llmbenchmark.ai/compare/grok-4-20-vs-minimax-minimax-m2-7">LLMBenchmark.ai, Grok 4.20 vs MiniMax M2.7</a></p>