Model comparison

Granite 4.2 8B vs MiniMax-M2.5

Granite 4.2 8B is the stronger model overall, scoring 40.0 to 38.1 on the Noometry Index.

Last verified October 10, 2026. 11 shared benchmarks.

Granite 4.2 8B IBM

40.0

Rank #143 Confirmed

MiniMax-M2.5

MiniMax

38.1

Rank #181 Confirmed

Summary

They share 11 benchmarks with published results for both. Granite 4.2 8B scores higher in 2 categories and MiniMax-M2.5 in 5 categories; 6 gaps are clear of the uncertainty.
The widest gap is in coding, where MiniMax-M2.5 leads 47.9 to 39.9.
Granite 4.2 8B is cheaper at $0.06 / $0.25 per million input/output tokens, against $0.30 / $1.20 for MiniMax-M2.5.
MiniMax-M2.5 accepts more context: 205K tokens versus 131K.

Side by side

Granite 4.2 8B and MiniMax-M2.5 specifications
	Granite 4.2 8B	MiniMax-M2.5
Provider	IBM	MiniMax
Noometry Index	40.0	38.1
Released	—	2026-02-12
Weights	Open	Open
Context window	131K	205K
Max output	118K	131K
Input $ / M tokens	$0.06	$0.30
Output $ / M tokens	$0.25	$1.20
Results tracked	11	33

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

Category by category

Coding MiniMax-M2.5 leads

Granite 4.2 8B: 39.9 (#138), MiniMax-M2.5: 47.9 (#53)

Coding benchmarks
Benchmark	Granite 4.2 8B	MiniMax-M2.5
LMArena Coding	1379	1380
SWE-bench Verified (bash only)	—	75.8%
LMArena WebDev	—	1386
SWE-bench Multilingual	—	68.3%
ALE-Bench	—	618.17

Agentic & Tool Use Not comparable

Granite 4.2 8B: —, MiniMax-M2.5: 30.2 (#70)

Agentic & Tool Use benchmarks
Benchmark	Granite 4.2 8B	MiniMax-M2.5
Terminal-Bench	—	42.7%
Vending-Bench 2	—	-23.16

Reasoning Granite 4.2 8B leads

Granite 4.2 8B: 25.5 (#130), MiniMax-M2.5: 17.6 (#281)

Reasoning benchmarks
Benchmark	Granite 4.2 8B	MiniMax-M2.5
LMArena Hard Prompts	1327	1372
ARC-AGI-2	—	4.9%
Kagi LLM Benchmark	—	55.2%
NYT Connections (extended)	—	16.8%
ARC-AGI-1	—	63.7%
Epoch Capabilities Index	—	146.68

Math Not comparable

Granite 4.2 8B: —, MiniMax-M2.5: 25.8 (#242)

Math benchmarks
Benchmark	Granite 4.2 8B	MiniMax-M2.5
ProofBench	—	4%
LMArena Math	—	1378

Knowledge Too close to call

Granite 4.2 8B: 37.7 (#144), MiniMax-M2.5: 38.6 (#132)

Knowledge benchmarks
Benchmark	Granite 4.2 8B	MiniMax-M2.5
LMArena Expert	1380	1378
Vectara Hallucination Rate	—	9.1%

Multilingual MiniMax-M2.5 leads

Granite 4.2 8B: 44.5 (#173), MiniMax-M2.5: 47.3 (#147)

Multilingual benchmarks
Benchmark	Granite 4.2 8B	MiniMax-M2.5
LMArena Non-English	1299	1338
LMArena Chinese	1361	1392
LMArena Russian	1283	1359
LMArena French	—	1363
LMArena German	—	1361
LMArena Japanese	—	1169
LMArena Korean	—	1232
LMArena Spanish	—	1354

Instruction Following MiniMax-M2.5 leads

Granite 4.2 8B: 68.7 (#183), MiniMax-M2.5: 71.7 (#143)

Instruction Following benchmarks
Benchmark	Granite 4.2 8B	MiniMax-M2.5
LMArena Instruction Following	1298	1353

Long Context Granite 4.2 8B leads

Granite 4.2 8B: 40.2 (#154), MiniMax-M2.5: 37.4 (#211)

Long Context benchmarks
Benchmark	Granite 4.2 8B	MiniMax-M2.5
LMArena Longer Query	1323	1367
CL-bench	—	11.4%
CL-bench Life	—	6.3%

Writing & Preference MiniMax-M2.5 leads

Granite 4.2 8B: 49.8 (#183), MiniMax-M2.5: 54.1 (#149)

Writing & Preference benchmarks
Benchmark	Granite 4.2 8B	MiniMax-M2.5
LMArena Text	1319	1359
LMArena Creative Writing	1238	1331
LMArena Multi-Turn	1299	1364
EQ-Bench Creative Writing	—	1361

Frequently asked questions

Is Granite 4.2 8B better than MiniMax-M2.5?

Granite 4.2 8B is the stronger model overall, scoring 40.0 to 38.1 on the Noometry Index.

Which is cheaper, Granite 4.2 8B or MiniMax-M2.5?

Granite 4.2 8B is cheaper. It lists at $0.06 per million input tokens and $0.25 per million output tokens; MiniMax-M2.5 lists at $0.30 and $1.20.

Is Granite 4.2 8B or MiniMax-M2.5 better for coding?

MiniMax-M2.5 scores higher on coding benchmarks: 47.9 versus 39.9 in the Noometry coding category.

Which has the bigger context window?

MiniMax-M2.5 does, with 205K tokens against 131K.

How many benchmarks do Granite 4.2 8B and MiniMax-M2.5 share?

11 benchmarks have published results for both models. Granite 4.2 8B has 11 scored results on Noometry and MiniMax-M2.5 has 33.