Model comparison
MiniMax-M2 vs o3
o3 is the stronger model overall, scoring 47.4 to 36.9 on the Noometry Index. MiniMax-M2 costs 6.7× less per token, which makes it the better buy when o3's lead doesn't matter for your workload.
Last verified . 17 shared benchmarks.
MiniMaxSummary
- They share 17 benchmarks with published results for both. MiniMax-M2 scores higher in 0 categories and o3 in 9 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where o3 leads 54.5 to 36.3.
- The biggest single-benchmark swing is Kagi LLM Benchmark: 57.8% for MiniMax-M2 and 67.6% for o3.
- MiniMax-M2 is cheaper at $0.30 / $1.20 per million input/output tokens, against $2 / $8 for o3.
- MiniMax-M2 accepts more context: 205K tokens versus 200K.
- MiniMax-M2 has downloadable open weights; the other is API-only.
Side by side
| MiniMax-M2 | o3 | |
|---|---|---|
| Provider | MiniMax | OpenAI |
| Noometry Index | 36.9 | 47.4 |
| Released | 2025-10-27 | 2025-04-16 |
| Weights | Open | Proprietary |
| Context window | 205K | 200K |
| Max output | 131K | 100K |
| Input $ / M tokens | $0.30 | $2 |
| Output $ / M tokens | $1.20 | $8 |
| Results tracked | 21 | 63 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding o3 leads
MiniMax-M2: 39.1 (#151), o3: 47.3 (#57)
| Benchmark | MiniMax-M2 | o3 |
|---|---|---|
| SWE-bench Verified (bash only) | 61% | 58.4% |
| LMArena Coding | 1370 | 1408 |
| SWE-bench Verified | — | 62.3% |
| Aider Polyglot | — | 81.3% |
| LMArena WebDev | 1297 | — |
| GSO | — | 8.8% |
| WeirdML | — | 52.4% |
| CadEval | — | 74% |
| ALE-Bench | — | 933.55 |
Agentic & Tool Use o3 leads
MiniMax-M2: 24.9 (#102), o3: 34.3 (#42)
| Benchmark | MiniMax-M2 | o3 |
|---|---|---|
| Terminal-Bench | 30% | — |
| Berkeley Function Calling Leaderboard | — | 63% |
| GDPval | — | 30.8% |
| DeepResearch Bench | — | 45.2% |
| OSWorld | — | 23% |
| LMArena Search | — | 1144 |
| METR Time Horizons | — | 65.4% |
| Vending-Bench 2 | 160.6 | — |
Reasoning o3 leads
MiniMax-M2: 18.6 (#256), o3: 32.1 (#66)
| Benchmark | MiniMax-M2 | o3 |
|---|---|---|
| Kagi LLM Benchmark | 57.8% | 67.6% |
| LMArena Hard Prompts | 1357 | 1402 |
| ARC-AGI-2 | — | 6.5% |
| SimpleBench | — | 53.1% |
| NYT Connections (extended) | 14.8% | — |
| ARC-AGI-1 | — | 60.8% |
| CritPt | — | 1.4% |
| Chess Puzzles | — | 38% |
| EnigmaEval | — | 13.1% |
| Mystery Game Puzzles | — | 29% |
| DTBench | — | 84.8% |
| LMCA | — | 39.7% |
| Epoch Capabilities Index | — | 146.88 |
| ForecastBench | — | 62.5 |
Math o3 leads
MiniMax-M2: 34.6 (#158), o3: 48.9 (#49)
| Benchmark | MiniMax-M2 | o3 |
|---|---|---|
| LMArena Math | 1350 | 1426 |
| FrontierMath (Tiers 1-3) | — | 33.3% |
| OTIS Mock AIME 2024-2025 | — | 84.4% |
| Omni-MATH | — | 71.4% |
| MATH Level 5 | — | 97.8% |
| FrontierMath (Feb 2025 set) | — | 18.7% |
| FrontierMath Tier 4 (v1) | — | 2.1% |
Knowledge o3 leads
MiniMax-M2: 36.3 (#160), o3: 54.5 (#47)
| Benchmark | MiniMax-M2 | o3 |
|---|---|---|
| LMArena Expert | 1336 | 1403 |
| GPQA Diamond | — | 81.8% |
| Humanity's Last Exam | — | 20.3% |
| SimpleQA Verified | — | 49.4% |
| MMLU-Pro | — | 85.9% |
| Confabulations | — | 14.4% |
| GPQA (HELM) | — | 75.3% |
Multimodal Not comparable
MiniMax-M2: —, o3: 40.7 (#35)
| Benchmark | MiniMax-M2 | o3 |
|---|---|---|
| LMArena Vision | — | 1214 |
| GeoBench | — | 74% |
| VPCT | — | 52% |
Multilingual o3 leads
MiniMax-M2: 45.6 (#165), o3: 51.9 (#99)
| Benchmark | MiniMax-M2 | o3 |
|---|---|---|
| LMArena Non-English | 1314 | 1402 |
| LMArena Chinese | 1367 | 1438 |
| LMArena French | 1336 | 1430 |
| LMArena German | 1366 | 1420 |
| LMArena Russian | 1332 | 1407 |
| LMArena Spanish | 1326 | 1395 |
| LMArena Japanese | — | 1404 |
| LMArena Korean | — | 1371 |
Instruction Following o3 leads
MiniMax-M2: 70.4 (#160), o3: 73.0 (#122)
| Benchmark | MiniMax-M2 | o3 |
|---|---|---|
| LMArena Instruction Following | 1329 | 1368 |
| IFEval | — | 86.9% |
Long Context o3 leads
MiniMax-M2: 40.5 (#148), o3: 53.2 (#6)
| Benchmark | MiniMax-M2 | o3 |
|---|---|---|
| LMArena Longer Query | 1332 | 1372 |
| Fiction.LiveBench | — | 88.9% |
| CL-bench | — | 17.8% |
Writing & Preference o3 leads
MiniMax-M2: 53.3 (#157), o3: 63.7 (#62)
| Benchmark | MiniMax-M2 | o3 |
|---|---|---|
| LMArena Text | 1340 | 1410 |
| LMArena Creative Writing | 1288 | 1359 |
| LMArena Multi-Turn | 1362 | 1405 |
| Short-Story Creative Writing | — | 83.9% |
| EQ-Bench Creative Writing | — | 1676 |
| WildBench | — | 86.1% |
Frequently asked questions
Is MiniMax-M2 better than o3?
o3 is the stronger model overall, scoring 47.4 to 36.9 on the Noometry Index. MiniMax-M2 costs 6.7× less per token, which makes it the better buy when o3's lead doesn't matter for your workload.
Which is cheaper, MiniMax-M2 or o3?
MiniMax-M2 is cheaper. It lists at $0.30 per million input tokens and $1.20 per million output tokens; o3 lists at $2 and $8.
Is MiniMax-M2 or o3 better for coding?
o3 scores higher on coding benchmarks: 47.3 versus 39.1 in the Noometry coding category.
Which has the bigger context window?
MiniMax-M2 does, with 205K tokens against 200K.
How many benchmarks do MiniMax-M2 and o3 share?
17 benchmarks have published results for both models. MiniMax-M2 has 21 scored results on Noometry and o3 has 63.