Noometry

Model comparison

GPT-4 vs MiniMax-M2.5

MiniMax-M2.5 is the stronger model overall, scoring 38.3 to 29.1 on the Noometry Index.

Last verified . 19 shared benchmarks.

GPT-4 OpenAI

29.1

Rank #316 Confirmed

MiniMax-M2.5 MiniMax

38.3

Rank #188 Confirmed

Summary

  • They share 19 benchmarks with published results for both. GPT-4 scores higher in 2 categories and MiniMax-M2.5 in 6 categories; 6 gaps are clear of the uncertainty.
  • The widest gap is in knowledge, where MiniMax-M2.5 leads 39.2 to 18.4.
  • MiniMax-M2.5 is cheaper at $0.30 / $1.20 per million input/output tokens, against $30 / $60 for GPT-4.
  • MiniMax-M2.5 accepts more context: 205K tokens versus 8K.
  • MiniMax-M2.5 has downloadable open weights; the other is API-only.

Side by side

GPT-4MiniMax-M2.5
ProviderOpenAIMiniMax
Noometry Index29.138.3
Released2023-03-142026-02-12
WeightsProprietaryOpen
Context window8K205K
Max output8K131K
Input $ / M tokens$30$0.30
Output $ / M tokens$60$1.20
Results tracked3833

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

Category by category

Coding MiniMax-M2.5 leads

GPT-4: 31.6 (#283), MiniMax-M2.5: 48.1 (#58)

BenchmarkGPT-4MiniMax-M2.5
LMArena Coding12541381
SWE-bench Verified (bash only)—75.8%
LMArena WebDev—1387
SWE-bench Multilingual—68.3%
WeirdML12.4%—
BigCodeBench Instruct46%—
BigCodeBench Complete57.2%—
ALE-Bench—618.17
HumanEval+79.3%—

Agentic & Tool Use Not comparable

GPT-4: —, MiniMax-M2.5: 30.4 (#77)

BenchmarkGPT-4MiniMax-M2.5
Terminal-Bench—42.7%
METR Time Horizons36.1%—
Vending-Bench 2—-23.16

Reasoning Too close to call

GPT-4: 17.8 (#289), MiniMax-M2.5: 17.5 (#292)

BenchmarkGPT-4MiniMax-M2.5
LMArena Hard Prompts12411372
Epoch Capabilities Index125.89146.68
ARC-AGI-2—4.9%
Kagi LLM Benchmark—55.2%
NYT Connections (extended)—16.8%
ARC-AGI-1—63.7%
Chess Puzzles4%—
Mystery Game Puzzles12%—
DTBench62.7%—
LMCA17.1%—
BIG-Bench Hard75.1%—
ForecastBench57.8—
HellaSwag95.3%—
WinoGrande87.5%—

Math MiniMax-M2.5 leads

GPT-4: 10.8 (#309), MiniMax-M2.5: 26.9 (#253)

BenchmarkGPT-4MiniMax-M2.5
LMArena Math12691378
OTIS Mock AIME 2024-20251.1%—
ProofBench—4%
MATH Level 523%—
GSM8K92%—

Knowledge MiniMax-M2.5 leads

GPT-4: 18.4 (#282), MiniMax-M2.5: 39.2 (#135)

BenchmarkGPT-4MiniMax-M2.5
LMArena Expert12111379
GPQA Diamond35.7%—
Vectara Hallucination Rate—9.1%
MMLU86.4%—
TriviaQA84.8%—

Multilingual MiniMax-M2.5 leads

GPT-4: 40.6 (#215), MiniMax-M2.5: 47.1 (#152)

BenchmarkGPT-4MiniMax-M2.5
LMArena Non-English12461338
LMArena Chinese12421393
LMArena French12831362
LMArena German12511362
LMArena Japanese12091171
LMArena Korean11841232
LMArena Russian12511358
LMArena Spanish12611354

Instruction Following MiniMax-M2.5 leads

GPT-4: 65.3 (#222), MiniMax-M2.5: 71.5 (#148)

BenchmarkGPT-4MiniMax-M2.5
LMArena Instruction Following12411353

Long Context Too close to call

GPT-4: 37.7 (#212), MiniMax-M2.5: 37.5 (#216)

BenchmarkGPT-4MiniMax-M2.5
LMArena Longer Query12441366
CL-bench—11.4%
CL-bench Life—6.3%

Writing & Preference MiniMax-M2.5 leads

GPT-4: 34.9 (#268), MiniMax-M2.5: 53.9 (#153)

BenchmarkGPT-4MiniMax-M2.5
LMArena Text12631359
LMArena Creative Writing12441331
EQ-Bench Creative Writing7521361
LMArena Multi-Turn12571364

Frequently asked questions

Is GPT-4 better than MiniMax-M2.5?

MiniMax-M2.5 is the stronger model overall, scoring 38.3 to 29.1 on the Noometry Index.

Which is cheaper, GPT-4 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; GPT-4 lists at $30 and $60.

Is GPT-4 or MiniMax-M2.5 better for coding?

MiniMax-M2.5 scores higher on coding benchmarks: 48.1 versus 31.6 in the Noometry coding category.

Which has the bigger context window?

MiniMax-M2.5 does, with 205K tokens against 8K.

How many benchmarks do GPT-4 and MiniMax-M2.5 share?

19 benchmarks have published results for both models. GPT-4 has 38 scored results on Noometry and MiniMax-M2.5 has 33.

Related comparisons

Go deeper