Model comparison

Grok 4.3 vs MiniMax-M2.5

Grok 4.3 is the stronger model overall, scoring 43.8 to 38.3 on the Noometry Index. MiniMax-M2.5 costs 3.0× less per token, which makes it the better buy when Grok 4.3's lead doesn't matter for your workload.

Last verified October 11, 2026. 23 shared benchmarks.

Grok 4.3 xAI

43.8

Rank #86 Confirmed

MiniMax-M2.5

MiniMax

38.3

Rank #188 Confirmed

Summary

They share 23 benchmarks with published results for both. Grok 4.3 scores higher in 7 categories and MiniMax-M2.5 in 2 categories; 8 gaps are clear of the uncertainty.
The widest gap is in math, where Grok 4.3 leads 46.0 to 26.9.
The biggest single-benchmark swing is NYT Connections (extended): 55.2% for Grok 4.3 and 16.8% for MiniMax-M2.5.
MiniMax-M2.5 is cheaper at $0.30 / $1.20 per million input/output tokens, against $1.25 / $2.50 for Grok 4.3.
Grok 4.3 accepts more context: 1M tokens versus 205K.
MiniMax-M2.5 has downloadable open weights; the other is API-only.

Side by side

Grok 4.3 and MiniMax-M2.5 specifications
	Grok 4.3	MiniMax-M2.5
Provider	xAI	MiniMax
Noometry Index	43.8	38.3
Released	2026-04-17	2026-02-12
Weights	Proprietary	Open
Context window	1M	205K
Max output	30K	131K
Input $ / M tokens	$1.25	$0.30
Output $ / M tokens	$2.50	$1.20
Results tracked	40	33

Sponsored placements are available on pages like this one. Advertise on Noometry

Category by category

Coding MiniMax-M2.5 leads

Grok 4.3: 41.6 (#121), MiniMax-M2.5: 48.1 (#58)

Coding benchmarks
Benchmark	Grok 4.3	MiniMax-M2.5
LMArena WebDev	1357	1387
LMArena Coding	1415	1381
ALE-Bench	944.17	618.17
SWE-bench Verified (bash only)	—	75.8%
SWE-bench Multilingual	—	68.3%
SciCode	47.3%	—
WeirdML	49.9%	—

Agentic & Tool Use MiniMax-M2.5 leads

Grok 4.3: 27.7 (#99), MiniMax-M2.5: 30.4 (#77)

Agentic & Tool Use benchmarks
Benchmark	Grok 4.3	MiniMax-M2.5
Vending-Bench 2	35.26	-23.16
Terminal-Bench	—	42.7%
GDP.pdf	8%	—
LMArena Search	1165	—

Reasoning Grok 4.3 leads

Grok 4.3: 35.9 (#68), MiniMax-M2.5: 17.5 (#292)

Reasoning benchmarks
Benchmark	Grok 4.3	MiniMax-M2.5
NYT Connections (extended)	55.2%	16.8%
LMArena Hard Prompts	1396	1372
Epoch Capabilities Index	149.16	146.68
ARC-AGI-2	—	4.9%
Kagi LLM Benchmark	—	55.2%
ARC-AGI-1	—	63.7%
CritPt	8%	—
Chess Puzzles	25%	—
DTBench	90.7%	—
LMCA	38.3%	—
ForecastBench	60.3	—

Math Grok 4.3 leads

Grok 4.3: 46.0 (#74), MiniMax-M2.5: 26.9 (#253)

Math benchmarks
Benchmark	Grok 4.3	MiniMax-M2.5
ProofBench	11%	4%
LMArena Math	1388	1378
FrontierMath (Tiers 1-3)	42.8%	—
FrontierMath Tier 4	14.6%	—
OTIS Mock AIME 2024-2025	93.3%	—

Knowledge Grok 4.3 leads

Grok 4.3: 52.5 (#62), MiniMax-M2.5: 39.2 (#135)

Knowledge benchmarks
Benchmark	Grok 4.3	MiniMax-M2.5
LMArena Expert	1385	1379
GPQA Diamond	88.8%	—
SimpleQA Verified	33.2%	—
Vectara Hallucination Rate	—	9.1%

Multimodal Not comparable

Grok 4.3: 31.6 (#104), MiniMax-M2.5: —

Multimodal benchmarks
Benchmark	Grok 4.3	MiniMax-M2.5
LMArena Vision	1229	—
Blueprint-Bench 2	0%	—

Multilingual Grok 4.3 leads

Grok 4.3: 50.5 (#120), MiniMax-M2.5: 47.1 (#152)

Multilingual benchmarks
Benchmark	Grok 4.3	MiniMax-M2.5
LMArena Non-English	1385	1338
LMArena Chinese	1422	1393
LMArena French	1412	1362
LMArena German	1395	1362
LMArena Japanese	1379	1171
LMArena Korean	1356	1232
LMArena Russian	1399	1358
LMArena Spanish	1398	1354

Instruction Following Too close to call

Grok 4.3: 72.1 (#140), MiniMax-M2.5: 71.5 (#148)

Instruction Following benchmarks
Benchmark	Grok 4.3	MiniMax-M2.5
LMArena Instruction Following	1366	1353

Long Context Grok 4.3 leads

Grok 4.3: 42.5 (#123), MiniMax-M2.5: 37.5 (#216)

Long Context benchmarks
Benchmark	Grok 4.3	MiniMax-M2.5
LMArena Longer Query	1393	1366
CL-bench	—	11.4%
CL-bench Life	—	6.3%

Writing & Preference Grok 4.3 leads

Grok 4.3: 58.5 (#118), MiniMax-M2.5: 53.9 (#153)

Writing & Preference benchmarks
Benchmark	Grok 4.3	MiniMax-M2.5
LMArena Text	1397	1359
LMArena Creative Writing	1380	1331
LMArena Multi-Turn	1406	1364
EQ-Bench Creative Writing	—	1361
EQ-Bench 4	1075	—

Frequently asked questions

Is Grok 4.3 better than MiniMax-M2.5?

Which is cheaper, Grok 4.3 or MiniMax-M2.5?

MiniMax-M2.5 is cheaper. It lists at $0.30 per million input tokens and $1.20 per million output tokens; Grok 4.3 lists at $1.25 and $2.50.

Is Grok 4.3 or MiniMax-M2.5 better for coding?

MiniMax-M2.5 scores higher on coding benchmarks: 48.1 versus 41.6 in the Noometry coding category.

Which has the bigger context window?

Grok 4.3 does, with 1M tokens against 205K.

How many benchmarks do Grok 4.3 and MiniMax-M2.5 share?

23 benchmarks have published results for both models. Grok 4.3 has 40 scored results on Noometry and MiniMax-M2.5 has 33.