Model comparison
GPT-4o vs MiniMax-M2.1
MiniMax-M2.1 is the stronger model overall, scoring 38.4 to 28.5 on the Noometry Index.
Last verified . 18 shared benchmarks.
MiniMaxSummary
- They share 18 benchmarks with published results for both. GPT-4o scores higher in 0 categories and MiniMax-M2.1 in 9 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in math, where MiniMax-M2.1 leads 35.6 to 10.2.
- MiniMax-M2.1 is cheaper at $0.30 / $1.20 per million input/output tokens, against $2.50 / $10 for GPT-4o.
- MiniMax-M2.1 accepts more context: 205K tokens versus 128K.
- MiniMax-M2.1 has downloadable open weights; the other is API-only.
Side by side
| GPT-4o | MiniMax-M2.1 | |
|---|---|---|
| Provider | OpenAI | MiniMax |
| Noometry Index | 28.5 | 38.4 |
| Released | 2024-05-13 | 2025-12-23 |
| Weights | Proprietary | Open |
| Context window | 128K | 205K |
| Max output | 16K | 131K |
| Input $ / M tokens | $2.50 | $0.30 |
| Output $ / M tokens | $10 | $1.20 |
| Results tracked | 72 | 22 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding MiniMax-M2.1 leads
GPT-4o: 25.1 (#323), MiniMax-M2.1: 40.1 (#132)
| Benchmark | GPT-4o | MiniMax-M2.1 |
|---|---|---|
| LMArena Coding | 1297 | 1421 |
| SWE-bench Verified | 31% | — |
| SWE-bench Verified (bash only) | 21.6% | — |
| Aider Polyglot | 45.3% | — |
| LMArena WebDev | — | 1384 |
| GSO | 0% | — |
| WeirdML | 25.1% | — |
| BigCodeBench Instruct | 51.1% | — |
| LiveBench Coding | 51.4% | — |
| BigCodeBench Complete | 61.1% | — |
| CadEval | 26% | — |
| ALE-Bench | — | 623.83 |
| HumanEval+ | 87.2% | — |
| MBPP+ | 72.2% | — |
Agentic & Tool Use MiniMax-M2.1 leads
GPT-4o: 20.8 (#133), MiniMax-M2.1: 27.7 (#91)
| Benchmark | GPT-4o | MiniMax-M2.1 |
|---|---|---|
| Terminal-Bench | — | 36.6% |
| GDPval | 9.9% | — |
| TheAgentCompany | 8.6% | — |
| Cybench | 12.5% | — |
| BALROG | 32.3% | — |
| LMArena Search | 1006 | — |
| METR Time Horizons | 40.8% | — |
Reasoning MiniMax-M2.1 leads
GPT-4o: 9.4 (#338), MiniMax-M2.1: 15.8 (#301)
| Benchmark | GPT-4o | MiniMax-M2.1 |
|---|---|---|
| LMArena Hard Prompts | 1281 | 1410 |
| ARC-AGI-2 | 0% | — |
| SimpleBench | 17.8% | — |
| NYT Connections (extended) | — | 11.2% |
| ARC-AGI-1 | 4.5% | — |
| CritPt | 0% | — |
| Chess Puzzles | 13% | — |
| EnigmaEval | 0.8% | — |
| LiveBench Reasoning | 55.8% | — |
| DTBench | 64.5% | — |
| LiveBench Data Analysis | 60.9% | — |
| LMCA | 16.6% | — |
| Epoch Capabilities Index | 128.97 | — |
| ForecastBench | 57.7 | — |
| LiveBench | 55.3% | — |
Math MiniMax-M2.1 leads
GPT-4o: 10.2 (#303), MiniMax-M2.1: 35.6 (#142)
| Benchmark | GPT-4o | MiniMax-M2.1 |
|---|---|---|
| LMArena Math | 1285 | 1396 |
| FrontierMath (Tiers 1-3) | 0.4% | — |
| OTIS Mock AIME 2024-2025 | 6.4% | — |
| Omni-MATH | 29.3% | — |
| LiveBench Math | 49.5% | — |
| MATH Level 5 | 53.3% | — |
| FrontierMath (Feb 2025 set) | 0.3% | — |
Knowledge MiniMax-M2.1 leads
GPT-4o: 28.8 (#235), MiniMax-M2.1: 37.7 (#143)
| Benchmark | GPT-4o | MiniMax-M2.1 |
|---|---|---|
| Vectara Hallucination Rate | 9.6% | 11.8% |
| LMArena Expert | 1250 | 1432 |
| GPQA Diamond | 49.2% | — |
| Humanity's Last Exam | 2.7% | — |
| SimpleQA Verified | 26% | — |
| MMLU-Pro | 71.3% | — |
| Confabulations | 15.3% | — |
| GPQA (HELM) | 52% | — |
| MMLU | 88.1% | — |
Multimodal Not comparable
GPT-4o: 33.9 (#85), MiniMax-M2.1: —
| Benchmark | GPT-4o | MiniMax-M2.1 |
|---|---|---|
| LMArena Vision | 1137 | — |
| Video-MME | 71.9% | — |
| GeoBench | 71% | — |
| VPCT | 40% | — |
| ScienceQA | 88.5% | — |
Multilingual MiniMax-M2.1 leads
GPT-4o: 43.4 (#181), MiniMax-M2.1: 50.2 (#123)
| Benchmark | GPT-4o | MiniMax-M2.1 |
|---|---|---|
| LMArena Non-English | 1283 | 1378 |
| LMArena Chinese | 1277 | 1430 |
| LMArena French | 1303 | 1405 |
| LMArena German | 1282 | 1380 |
| LMArena Japanese | 1257 | 1287 |
| LMArena Korean | 1234 | 1297 |
| LMArena Russian | 1286 | 1386 |
| LMArena Spanish | 1292 | 1396 |
Instruction Following MiniMax-M2.1 leads
GPT-4o: 66.8 (#202), MiniMax-M2.1: 74.0 (#108)
| Benchmark | GPT-4o | MiniMax-M2.1 |
|---|---|---|
| LMArena Instruction Following | 1278 | 1399 |
| LiveBench Instruction Following | 68.6% | — |
| IFEval | 81.7% | — |
Long Context MiniMax-M2.1 leads
GPT-4o: 39.4 (#174), MiniMax-M2.1: 43.2 (#98)
| Benchmark | GPT-4o | MiniMax-M2.1 |
|---|---|---|
| LMArena Longer Query | 1289 | 1416 |
| Fiction.LiveBench | 66.7% | — |
Writing & Preference MiniMax-M2.1 leads
GPT-4o: 52.8 (#160), MiniMax-M2.1: 58.5 (#114)
| Benchmark | GPT-4o | MiniMax-M2.1 |
|---|---|---|
| LMArena Text | 1300 | 1392 |
| LMArena Creative Writing | 1291 | 1362 |
| LMArena Multi-Turn | 1302 | 1396 |
| Short-Story Creative Writing | 81.8% | — |
| WildBench | 82.8% | — |
| LiveBench Language | 47.6% | — |
Frequently asked questions
Is GPT-4o better than MiniMax-M2.1?
MiniMax-M2.1 is the stronger model overall, scoring 38.4 to 28.5 on the Noometry Index.
Which is cheaper, GPT-4o 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-4o lists at $2.50 and $10.
Is GPT-4o or MiniMax-M2.1 better for coding?
MiniMax-M2.1 scores higher on coding benchmarks: 40.1 versus 25.1 in the Noometry coding category.
Which has the bigger context window?
MiniMax-M2.1 does, with 205K tokens against 128K.
How many benchmarks do GPT-4o and MiniMax-M2.1 share?
18 benchmarks have published results for both models. GPT-4o has 72 scored results on Noometry and MiniMax-M2.1 has 22.