Model comparison
GPT-4.1 vs MiniMax-M2.1
MiniMax-M2.1 is the stronger model overall, scoring 38.4 to 35.7 on the Noometry Index.
Last verified . 19 shared benchmarks.
MiniMaxSummary
- They share 19 benchmarks with published results for both. GPT-4.1 scores higher in 1 category and MiniMax-M2.1 in 8 categories; 6 gaps are clear of the uncertainty.
- The widest gap is in math, where MiniMax-M2.1 leads 35.6 to 21.4.
- The biggest single-benchmark swing is Vectara Hallucination Rate: 5.6% for GPT-4.1 and 11.8% for MiniMax-M2.1.
- MiniMax-M2.1 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 205K.
- MiniMax-M2.1 has downloadable open weights; the other is API-only.
Side by side
| GPT-4.1 | MiniMax-M2.1 | |
|---|---|---|
| Provider | OpenAI | MiniMax |
| Noometry Index | 35.7 | 38.4 |
| Released | 2025-04-14 | 2025-12-23 |
| Weights | Proprietary | Open |
| Context window | 1.05M | 205K |
| Max output | 33K | 131K |
| Input $ / M tokens | $2 | $0.30 |
| Output $ / M tokens | $8 | $1.20 |
| Results tracked | 52 | 22 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding MiniMax-M2.1 leads
GPT-4.1: 34.2 (#228), MiniMax-M2.1: 40.1 (#132)
| Benchmark | GPT-4.1 | MiniMax-M2.1 |
|---|---|---|
| LMArena Coding | 1391 | 1421 |
| ALE-Bench | 558.1 | 623.83 |
| SWE-bench Verified | 48.5% | — |
| SWE-bench Verified (bash only) | 39.6% | — |
| Aider Polyglot | 52.4% | — |
| LMArena WebDev | — | 1384 |
| WeirdML | 39% | — |
| CadEval | 42% | — |
Agentic & Tool Use GPT-4.1 leads
GPT-4.1: 34.4 (#40), MiniMax-M2.1: 27.7 (#91)
| Benchmark | GPT-4.1 | MiniMax-M2.1 |
|---|---|---|
| Terminal-Bench | — | 36.6% |
| Berkeley Function Calling Leaderboard | 54% | — |
Reasoning MiniMax-M2.1 leads
GPT-4.1: 11.7 (#334), MiniMax-M2.1: 15.8 (#301)
| Benchmark | GPT-4.1 | MiniMax-M2.1 |
|---|---|---|
| LMArena Hard Prompts | 1384 | 1410 |
| ARC-AGI-2 | 0.4% | — |
| SimpleBench | 27% | — |
| Kagi LLM Benchmark | 52.3% | — |
| NYT Connections (extended) | — | 11.2% |
| ARC-AGI-1 | 5.5% | — |
| Chess Puzzles | 6% | — |
| EnigmaEval | 2.2% | — |
| DTBench | 68.3% | — |
| LMCA | 25.6% | — |
| Epoch Capabilities Index | 136.79 | — |
| ForecastBench | 61.5 | — |
Math MiniMax-M2.1 leads
GPT-4.1: 21.4 (#269), MiniMax-M2.1: 35.6 (#142)
| Benchmark | GPT-4.1 | MiniMax-M2.1 |
|---|---|---|
| LMArena Math | 1369 | 1396 |
| FrontierMath (Tiers 1-3) | 6% | — |
| OTIS Mock AIME 2024-2025 | 38.3% | — |
| Omni-MATH | 47.1% | — |
| MATH Level 5 | 83% | — |
| FrontierMath (Feb 2025 set) | 5.5% | — |
| FrontierMath Tier 4 (v1) | 0% | — |
Knowledge Too close to call
GPT-4.1: 37.0 (#152), MiniMax-M2.1: 37.7 (#143)
| Benchmark | GPT-4.1 | MiniMax-M2.1 |
|---|---|---|
| Vectara Hallucination Rate | 5.6% | 11.8% |
| LMArena Expert | 1364 | 1432 |
| GPQA Diamond | 66.9% | — |
| Humanity's Last Exam | 5.4% | — |
| SimpleQA Verified | 31.1% | — |
| MMLU-Pro | 81.1% | — |
| GPQA (HELM) | 65.9% | — |
Multimodal Not comparable
GPT-4.1: 37.5 (#66), MiniMax-M2.1: —
| Benchmark | GPT-4.1 | MiniMax-M2.1 |
|---|---|---|
| LMArena Vision | 1210 | — |
| GeoBench | 72% | — |
Multilingual Too close to call
GPT-4.1: 49.6 (#128), MiniMax-M2.1: 50.2 (#123)
| Benchmark | GPT-4.1 | MiniMax-M2.1 |
|---|---|---|
| LMArena Non-English | 1370 | 1378 |
| LMArena Chinese | 1382 | 1430 |
| LMArena French | 1383 | 1405 |
| LMArena German | 1380 | 1380 |
| LMArena Japanese | 1320 | 1287 |
| LMArena Korean | 1339 | 1297 |
| LMArena Russian | 1377 | 1386 |
| LMArena Spanish | 1375 | 1396 |
Instruction Following MiniMax-M2.1 leads
GPT-4.1: 71.5 (#148), MiniMax-M2.1: 74.0 (#108)
| Benchmark | GPT-4.1 | MiniMax-M2.1 |
|---|---|---|
| LMArena Instruction Following | 1367 | 1399 |
| IFEval | 83.8% | — |
Long Context MiniMax-M2.1 leads
GPT-4.1: 40.0 (#158), MiniMax-M2.1: 43.2 (#98)
| Benchmark | GPT-4.1 | MiniMax-M2.1 |
|---|---|---|
| LMArena Longer Query | 1385 | 1416 |
| Fiction.LiveBench | 63.9% | — |
Writing & Preference Too close to call
GPT-4.1: 57.8 (#120), MiniMax-M2.1: 58.5 (#114)
| Benchmark | GPT-4.1 | MiniMax-M2.1 |
|---|---|---|
| LMArena Text | 1383 | 1392 |
| LMArena Creative Writing | 1363 | 1362 |
| LMArena Multi-Turn | 1398 | 1396 |
| EQ-Bench Creative Writing | 1420 | — |
| WildBench | 85.4% | — |
Frequently asked questions
Is GPT-4.1 better than MiniMax-M2.1?
MiniMax-M2.1 is the stronger model overall, scoring 38.4 to 35.7 on the Noometry Index.
Which is cheaper, GPT-4.1 or MiniMax-M2.1?
MiniMax-M2.1 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-M2.1 better for coding?
MiniMax-M2.1 scores higher on coding benchmarks: 40.1 versus 34.2 in the Noometry coding category.
Which has the bigger context window?
GPT-4.1 does, with 1.05M tokens against 205K.
How many benchmarks do GPT-4.1 and MiniMax-M2.1 share?
19 benchmarks have published results for both models. GPT-4.1 has 52 scored results on Noometry and MiniMax-M2.1 has 22.