Model comparison

GPT-5.4 nano vs MiniMax-M2.5

GPT-5.4 nano is the stronger model overall, scoring 41.9 to 38.3 on the Noometry Index.

Last verified October 12, 2026. 24 shared benchmarks.

GPT-5.4 nano OpenAI

41.9

Rank #125 Confirmed

MiniMax-M2.5

MiniMax

38.3

Rank #188 Confirmed

Summary

They share 24 benchmarks with published results for both. GPT-5.4 nano scores higher in 7 categories and MiniMax-M2.5 in 1 category; 7 gaps are clear of the uncertainty.
The widest gap is in math, where GPT-5.4 nano leads 40.9 to 26.9.
The biggest single-benchmark swing is Kagi LLM Benchmark: 39.7% for GPT-5.4 nano and 55.2% for MiniMax-M2.5.
GPT-5.4 nano is cheaper at $0.20 / $1.25 per million input/output tokens, against $0.30 / $1.20 for MiniMax-M2.5.
GPT-5.4 nano accepts more context: 400K tokens versus 205K.
MiniMax-M2.5 has downloadable open weights; the other is API-only.

Side by side

GPT-5.4 nano and MiniMax-M2.5 specifications
	GPT-5.4 nano	MiniMax-M2.5
Provider	OpenAI	MiniMax
Noometry Index	41.9	38.3
Released	2026-03-17	2026-02-12
Weights	Proprietary	Open
Context window	400K	205K
Max output	128K	131K
Input $ / M tokens	$0.20	$0.30
Output $ / M tokens	$1.25	$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

GPT-5.4 nano: 43.6 (#84), MiniMax-M2.5: 48.1 (#58)

Coding benchmarks
Benchmark	GPT-5.4 nano	MiniMax-M2.5
LMArena Coding	1405	1381
ALE-Bench	1,005	618.17
SWE-bench Verified (bash only)	—	75.8%
LMArena WebDev	—	1387
SWE-bench Multilingual	—	68.3%
SciCode	46.9%	—
WeirdML	49.2%	—

Agentic & Tool Use Not comparable

GPT-5.4 nano: —, MiniMax-M2.5: 30.4 (#77)

Agentic & Tool Use benchmarks
Benchmark	GPT-5.4 nano	MiniMax-M2.5
Terminal-Bench	—	42.7%
Vending-Bench 2	—	-23.16

Reasoning GPT-5.4 nano leads

GPT-5.4 nano: 23.7 (#173), MiniMax-M2.5: 17.5 (#292)

Reasoning benchmarks
Benchmark	GPT-5.4 nano	MiniMax-M2.5
ARC-AGI-2	5.7%	4.9%
Kagi LLM Benchmark	39.7%	55.2%
ARC-AGI-1	51.5%	63.7%
LMArena Hard Prompts	1381	1372
Epoch Capabilities Index	145.81	146.68
NYT Connections (extended)	—	16.8%
CritPt	9.3%	—
Chess Puzzles	30%	—
Mystery Game Puzzles	9%	—
DTBench	80.3%	—
LMCA	36.9%	—
ForecastBench	57.3	—

Math GPT-5.4 nano leads

GPT-5.4 nano: 40.9 (#88), MiniMax-M2.5: 26.9 (#253)

Math benchmarks
Benchmark	GPT-5.4 nano	MiniMax-M2.5
ProofBench	5%	4%
LMArena Math	1406	1378
FrontierMath (Tiers 1-3)	44.9%	—
FrontierMath Tier 4	12.2%	—
OTIS Mock AIME 2024-2025	87.8%	—
FrontierMath (Feb 2025 set)	25.9%	—
FrontierMath Tier 4 (v1)	6.3%	—

Knowledge GPT-5.4 nano leads

GPT-5.4 nano: 41.9 (#103), MiniMax-M2.5: 39.2 (#135)

Knowledge benchmarks
Benchmark	GPT-5.4 nano	MiniMax-M2.5
Vectara Hallucination Rate	3.1%	9.1%
LMArena Expert	1396	1379
GPQA Diamond	78.5%	—
SimpleQA Verified	11.7%	—

Multimodal Not comparable

GPT-5.4 nano: 36.7 (#78), MiniMax-M2.5: —

Multimodal benchmarks
Benchmark	GPT-5.4 nano	MiniMax-M2.5
LMArena Vision	1196	—

Multilingual GPT-5.4 nano leads

GPT-5.4 nano: 48.6 (#140), MiniMax-M2.5: 47.1 (#152)

Multilingual benchmarks
Benchmark	GPT-5.4 nano	MiniMax-M2.5
LMArena Non-English	1359	1338
LMArena Chinese	1392	1393
LMArena French	1396	1362
LMArena German	1367	1362
LMArena Japanese	1343	1171
LMArena Korean	1320	1232
LMArena Russian	1363	1358
LMArena Spanish	1371	1354

Instruction Following Too close to call

GPT-5.4 nano: 71.9 (#144), MiniMax-M2.5: 71.5 (#148)

Instruction Following benchmarks
Benchmark	GPT-5.4 nano	MiniMax-M2.5
LMArena Instruction Following	1362	1353

Long Context GPT-5.4 nano leads

GPT-5.4 nano: 41.6 (#137), MiniMax-M2.5: 37.5 (#216)

Long Context benchmarks
Benchmark	GPT-5.4 nano	MiniMax-M2.5
LMArena Longer Query	1366	1366
CL-bench	—	11.4%
CL-bench Life	—	6.3%

Writing & Preference GPT-5.4 nano leads

GPT-5.4 nano: 55.7 (#142), MiniMax-M2.5: 53.9 (#153)

Writing & Preference benchmarks
Benchmark	GPT-5.4 nano	MiniMax-M2.5
LMArena Text	1372	1359
LMArena Creative Writing	1314	1331
LMArena Multi-Turn	1382	1364
EQ-Bench Creative Writing	—	1361

Frequently asked questions

Is GPT-5.4 nano better than MiniMax-M2.5?

GPT-5.4 nano is the stronger model overall, scoring 41.9 to 38.3 on the Noometry Index.

Which is cheaper, GPT-5.4 nano or MiniMax-M2.5?

GPT-5.4 nano is cheaper. It lists at $0.20 per million input tokens and $1.25 per million output tokens; MiniMax-M2.5 lists at $0.30 and $1.20.

Is GPT-5.4 nano or MiniMax-M2.5 better for coding?

MiniMax-M2.5 scores higher on coding benchmarks: 48.1 versus 43.6 in the Noometry coding category.

Which has the bigger context window?

GPT-5.4 nano does, with 400K tokens against 205K.

How many benchmarks do GPT-5.4 nano and MiniMax-M2.5 share?

24 benchmarks have published results for both models. GPT-5.4 nano has 40 scored results on Noometry and MiniMax-M2.5 has 33.