# GPT-4.1 vs MiniMax-M3

> MiniMax-M3 is the stronger model overall, scoring 43.5 to 35.7 on the Noometry Index.

- Canonical page: https://noometry.com/compare/gpt-4-1-vs-minimax-m3
- Last updated: 2026-10-10
- Shared benchmarks: 27

## Summary

- They share 27 benchmarks with published results for both. GPT-4.1 scores higher in 1 category and MiniMax-M3 in 9 categories; 10 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where MiniMax-M3 leads 58.4 to 37.0.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 38.3% for GPT-4.1 and 71.1% for MiniMax-M3.
- MiniMax-M3 is cheaper at $0.30 / $1.20 per million input/output tokens, against $2 / $8 for GPT-4.1.
- GPT-4.1 accepts more context: 1.05M tokens versus 1M.
- MiniMax-M3 has downloadable open weights; the other is API-only.

## Snapshot

| | GPT-4.1 | MiniMax-M3 |
|---|---|---|
| Provider | OpenAI | MiniMax |
| Noometry Index | 35.7 | 43.5 |
| Rank | 215 | 81 |
| Context | 1.05M | 1M |
| Input $/M | $2 | $0.30 |
| Output $/M | $8 | $1.20 |
| Weights | Proprietary | Open |

## Coding

- GPT-4.1: 34.2 (#228)
- MiniMax-M3: 41.5 (#111)

| Benchmark | GPT-4.1 | MiniMax-M3 |
|---|---|---|
| LMArena Coding | 1391 | 1470 |
| ALE-Bench | 558.1 | 640.02 |
| SWE-bench Verified | 48.5% | — |
| FrontierCode | — | 14.7% |
| SWE-bench Verified (bash only) | 39.6% | — |
| Aider Polyglot | 52.4% | — |
| LMArena WebDev | — | 1482 |
| SciCode | — | 47.1% |
| WeirdML | 39% | — |
| CadEval | 42% | — |

## Agentic & Tool Use

- GPT-4.1: 34.4 (#40)
- MiniMax-M3: 22.5 (#121)

| Benchmark | GPT-4.1 | MiniMax-M3 |
|---|---|---|
| APEX-Agents | — | 37.7% |
| Berkeley Function Calling Leaderboard | 54% | — |
| OSWorld 2.0 | — | 4.6% |
| GBAEval | — | 0.9% |
| Vending-Bench 2 | — | 2,158 |

## Reasoning

- GPT-4.1: 11.7 (#334)
- MiniMax-M3: 29.3 (#83)

| Benchmark | GPT-4.1 | MiniMax-M3 |
|---|---|---|
| SimpleBench | 27% | 45.8% |
| Chess Puzzles | 6% | 14% |
| LMArena Hard Prompts | 1384 | 1447 |
| DTBench | 68.3% | 78.9% |
| LMCA | 25.6% | 33.7% |
| Epoch Capabilities Index | 136.79 | 147 |
| ForecastBench | 61.5 | 61.4 |
| ARC-AGI-2 | 0.4% | — |
| Kagi LLM Benchmark | 52.3% | — |
| NYT Connections (extended) | — | 65.1% |
| ARC-AGI-1 | 5.5% | — |
| CritPt | — | 3.7% |
| EnigmaEval | 2.2% | — |
| Mystery Game Puzzles | — | 8% |
| Surface Evolver Bench | — | 55% |

## Math

- GPT-4.1: 21.4 (#269)
- MiniMax-M3: 39.1 (#83)

| Benchmark | GPT-4.1 | MiniMax-M3 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 38.3% | 71.1% |
| LMArena Math | 1369 | 1432 |
| FrontierMath (Tiers 1-3) | 6% | — |
| ProofBench | — | 18% |
| Omni-MATH | 47.1% | — |
| MATH Level 5 | 83% | — |
| FrontierMath (Feb 2025 set) | 5.5% | — |
| FrontierMath Tier 4 (v1) | 0% | — |

## Knowledge

- GPT-4.1: 37.0 (#152)
- MiniMax-M3: 58.4 (#32)

| Benchmark | GPT-4.1 | MiniMax-M3 |
|---|---|---|
| GPQA Diamond | 66.9% | 90.9% |
| LMArena Expert | 1364 | 1462 |
| Humanity's Last Exam | 5.4% | — |
| SimpleQA Verified | 31.1% | — |
| MMLU-Pro | 81.1% | — |
| Vectara Hallucination Rate | 5.6% | — |
| GPQA (HELM) | 65.9% | — |

## Multimodal

- GPT-4.1: 37.5 (#66)
- MiniMax-M3: 39.5 (#45)

| Benchmark | GPT-4.1 | MiniMax-M3 |
|---|---|---|
| LMArena Vision | 1210 | 1254 |
| GeoBench | 72% | — |
| LMArena Document | — | 1435 |

## Multilingual

- GPT-4.1: 49.6 (#128)
- MiniMax-M3: 53.2 (#70)

| Benchmark | GPT-4.1 | MiniMax-M3 |
|---|---|---|
| LMArena Non-English | 1370 | 1420 |
| LMArena Chinese | 1382 | 1465 |
| LMArena French | 1383 | 1445 |
| LMArena German | 1380 | 1424 |
| LMArena Japanese | 1320 | 1381 |
| LMArena Korean | 1339 | 1372 |
| LMArena Russian | 1377 | 1430 |
| LMArena Spanish | 1375 | 1431 |

## Instruction Following

- GPT-4.1: 71.5 (#148)
- MiniMax-M3: 75.7 (#56)

| Benchmark | GPT-4.1 | MiniMax-M3 |
|---|---|---|
| LMArena Instruction Following | 1367 | 1433 |
| IFEval | 83.8% | — |

## Long Context

- GPT-4.1: 40.0 (#158)
- MiniMax-M3: 44.2 (#68)

| Benchmark | GPT-4.1 | MiniMax-M3 |
|---|---|---|
| LMArena Longer Query | 1385 | 1445 |
| Fiction.LiveBench | 63.9% | — |

## Writing & Preference

- GPT-4.1: 57.8 (#120)
- MiniMax-M3: 62.3 (#79)

| Benchmark | GPT-4.1 | MiniMax-M3 |
|---|---|---|
| LMArena Text | 1383 | 1433 |
| LMArena Creative Writing | 1363 | 1405 |
| LMArena Multi-Turn | 1398 | 1441 |
| EQ-Bench Creative Writing | 1420 | — |
| WildBench | 85.4% | — |
| EQ-Bench 4 | — | 1150 |

## FAQ

### Is GPT-4.1 better than MiniMax-M3?

MiniMax-M3 is the stronger model overall, scoring 43.5 to 35.7 on the Noometry Index.

### Which is cheaper, GPT-4.1 or MiniMax-M3?

MiniMax-M3 is cheaper. It lists at $0.30 per million input tokens and $1.20 per million output tokens; GPT-4.1 lists at $2 and $8.

### Is GPT-4.1 or MiniMax-M3 better for coding?

MiniMax-M3 scores higher on coding benchmarks: 41.5 versus 34.2 in the Noometry coding category.

### Which has the bigger context window?

GPT-4.1 does, with 1.05M tokens against 1M.

### How many benchmarks do GPT-4.1 and MiniMax-M3 share?

27 benchmarks have published results for both models. GPT-4.1 has 52 scored results on Noometry and MiniMax-M3 has 41.