Model comparison

GPT-5.3 Codex vs MiniMax-M2

GPT-5.3 Codex is the stronger model overall, scoring 45.8 to 37.4 on the Noometry Index. MiniMax-M2 costs 9.2× less per token, which makes it the better buy when GPT-5.3 Codex's lead doesn't matter for your workload.

Last verified October 10, 2026. 3 shared benchmarks.

GPT-5.3 Codex OpenAI

45.8

Rank #69 Reported

MiniMax-M2

MiniMax

37.4

Rank #204 Confirmed

Summary

They share 3 benchmarks with published results for both. GPT-5.3 Codex scores higher in 2 categories and MiniMax-M2 in 0 categories; 2 gaps are clear of the uncertainty.
The widest gap is in agentic & tool use, where GPT-5.3 Codex leads 48.0 to 25.1.
The biggest single-benchmark swing is Terminal-Bench: 78.4% for GPT-5.3 Codex and 30% for MiniMax-M2.
MiniMax-M2 is cheaper at $0.30 / $1.20 per million input/output tokens, against $1.75 / $14 for GPT-5.3 Codex.
GPT-5.3 Codex accepts more context: 400K tokens versus 205K.
MiniMax-M2 has downloadable open weights; the other is API-only.

Side by side

GPT-5.3 Codex and MiniMax-M2 specifications
	GPT-5.3 Codex	MiniMax-M2
Provider	OpenAI	MiniMax
Noometry Index	45.8	37.4
Released	2026-02-05	2025-10-27
Weights	Proprietary	Open
Context window	400K	205K
Max output	128K	131K
Input $ / M tokens	$1.75	$0.30
Output $ / M tokens	$14	$1.20
Results tracked	8	21

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

Category by category

Coding GPT-5.3 Codex leads

GPT-5.3 Codex: 48.6 (#56), MiniMax-M2: 39.3 (#159)

Coding benchmarks
Benchmark	GPT-5.3 Codex	MiniMax-M2
LMArena WebDev	1409	1297
SWE-bench Verified	74.8%	—
SWE-bench Verified (bash only)	—	61%
WeirdML	79.3%	—
LMArena Coding	—	1370
ALE-Bench	1,655	—

Agentic & Tool Use GPT-5.3 Codex leads

GPT-5.3 Codex: 48.0 (#9), MiniMax-M2: 25.1 (#109)

Agentic & Tool Use benchmarks
Benchmark	GPT-5.3 Codex	MiniMax-M2
Terminal-Bench	78.4%	30%
Vending-Bench 2	5,940	160.6
METR Time Horizons	74.5%	—

Reasoning Not comparable

GPT-5.3 Codex: —, MiniMax-M2: 19.4 (#258)

Reasoning benchmarks
Benchmark	GPT-5.3 Codex	MiniMax-M2
Kagi LLM Benchmark	—	57.8%
NYT Connections (extended)	—	14.8%
LMArena Hard Prompts	—	1357
Epoch Capabilities Index	156.77	—

Math Not comparable

GPT-5.3 Codex: —, MiniMax-M2: 37.3 (#160)

Math benchmarks
Benchmark	GPT-5.3 Codex	MiniMax-M2
LMArena Math	—	1352

Knowledge Not comparable

GPT-5.3 Codex: —, MiniMax-M2: 37.0 (#163)

Knowledge benchmarks
Benchmark	GPT-5.3 Codex	MiniMax-M2
LMArena Expert	—	1337

Multilingual Not comparable

GPT-5.3 Codex: —, MiniMax-M2: 45.3 (#171)

Multilingual benchmarks
Benchmark	GPT-5.3 Codex	MiniMax-M2
LMArena Non-English	—	1313
LMArena Chinese	—	1366
LMArena French	—	1335
LMArena German	—	1355
LMArena Russian	—	1331
LMArena Spanish	—	1326

Instruction Following Not comparable

GPT-5.3 Codex: —, MiniMax-M2: 70.2 (#166)

Instruction Following benchmarks
Benchmark	GPT-5.3 Codex	MiniMax-M2
LMArena Instruction Following	—	1328

Long Context Not comparable

GPT-5.3 Codex: —, MiniMax-M2: 40.5 (#153)

Long Context benchmarks
Benchmark	GPT-5.3 Codex	MiniMax-M2
LMArena Longer Query	—	1331

Writing & Preference Not comparable

GPT-5.3 Codex: —, MiniMax-M2: 53.0 (#162)

Writing & Preference benchmarks
Benchmark	GPT-5.3 Codex	MiniMax-M2
LMArena Text	—	1340
LMArena Creative Writing	—	1286
LMArena Multi-Turn	—	1361

Frequently asked questions

Is GPT-5.3 Codex better than MiniMax-M2?

Which is cheaper, GPT-5.3 Codex or MiniMax-M2?

MiniMax-M2 is cheaper. It lists at $0.30 per million input tokens and $1.20 per million output tokens; GPT-5.3 Codex lists at $1.75 and $14.

Is GPT-5.3 Codex or MiniMax-M2 better for coding?

GPT-5.3 Codex scores higher on coding benchmarks: 48.6 versus 39.3 in the Noometry coding category.

Which has the bigger context window?

GPT-5.3 Codex does, with 400K tokens against 205K.

How many benchmarks do GPT-5.3 Codex and MiniMax-M2 share?

3 benchmarks have published results for both models. GPT-5.3 Codex has 8 scored results on Noometry and MiniMax-M2 has 21.