Model comparison
MiniMax-M3 vs Qwen2.5-Max
MiniMax-M3 is the stronger model overall, scoring 43.5 to 40.1 on the Noometry Index.
Last verified . 18 shared benchmarks.
MiniMax
Alibaba (Qwen)Summary
- They share 18 benchmarks with published results for both. MiniMax-M3 scores higher in 8 categories and Qwen2.5-Max in 0 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where MiniMax-M3 leads 58.4 to 34.8.
- MiniMax-M3 has downloadable open weights; the other is API-only.
Side by side
| MiniMax-M3 | Qwen2.5-Max | |
|---|---|---|
| Provider | MiniMax | Alibaba (Qwen) |
| Noometry Index | 43.5 | 40.1 |
| Released | 2026-06-01 | 2025-01-25 |
| Weights | Open | Proprietary |
| Context window | 1M | — |
| Max output | 512K | — |
| Input $ / M tokens | $0.30 | — |
| Output $ / M tokens | $1.20 | — |
| Results tracked | 41 | 27 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Too close to call
MiniMax-M3: 41.5 (#111), Qwen2.5-Max: 41.2 (#118)
| Benchmark | MiniMax-M3 | Qwen2.5-Max |
|---|---|---|
| LMArena Coding | 1470 | 1359 |
| FrontierCode | 14.7% | — |
| LMArena WebDev | 1482 | — |
| SciCode | 47.1% | — |
| LiveBench Coding | — | 64.4% |
| ALE-Bench | 640.02 | — |
Agentic & Tool Use Not comparable
MiniMax-M3: 22.5 (#121), Qwen2.5-Max: —
| Benchmark | MiniMax-M3 | Qwen2.5-Max |
|---|---|---|
| APEX-Agents | 37.7% | — |
| OSWorld 2.0 | 4.6% | — |
| GBAEval | 0.9% | — |
| Vending-Bench 2 | 2,158 | — |
Reasoning MiniMax-M3 leads
MiniMax-M3: 29.3 (#83), Qwen2.5-Max: 24.6 (#144)
| Benchmark | MiniMax-M3 | Qwen2.5-Max |
|---|---|---|
| LMArena Hard Prompts | 1447 | 1360 |
| Epoch Capabilities Index | 147 | 132.54 |
| SimpleBench | 45.8% | — |
| NYT Connections (extended) | 65.1% | — |
| CritPt | 3.7% | — |
| Chess Puzzles | 14% | — |
| LiveBench Reasoning | — | 51.4% |
| Mystery Game Puzzles | 8% | — |
| DTBench | 78.9% | — |
| LiveBench Data Analysis | — | 67.9% |
| LMCA | 33.7% | — |
| Surface Evolver Bench | 55% | — |
| ForecastBench | 61.4 | — |
| LiveBench | — | 62.3% |
Math MiniMax-M3 leads
MiniMax-M3: 39.1 (#83), Qwen2.5-Max: 34.3 (#162)
| Benchmark | MiniMax-M3 | Qwen2.5-Max |
|---|---|---|
| LMArena Math | 1432 | 1369 |
| OTIS Mock AIME 2024-2025 | 71.1% | — |
| ProofBench | 18% | — |
| LiveBench Math | — | 58.4% |
Knowledge MiniMax-M3 leads
MiniMax-M3: 58.4 (#32), Qwen2.5-Max: 34.8 (#181)
| Benchmark | MiniMax-M3 | Qwen2.5-Max |
|---|---|---|
| LMArena Expert | 1462 | 1337 |
| GPQA Diamond | 90.9% | — |
| Confabulations | — | 21.8% |
Multimodal Not comparable
MiniMax-M3: 39.5 (#45), Qwen2.5-Max: —
| Benchmark | MiniMax-M3 | Qwen2.5-Max |
|---|---|---|
| LMArena Vision | 1254 | — |
| LMArena Document | 1435 | — |
Multilingual MiniMax-M3 leads
MiniMax-M3: 53.2 (#70), Qwen2.5-Max: 48.3 (#140)
| Benchmark | MiniMax-M3 | Qwen2.5-Max |
|---|---|---|
| LMArena Non-English | 1420 | 1352 |
| LMArena Chinese | 1465 | 1382 |
| LMArena French | 1445 | 1396 |
| LMArena German | 1424 | 1350 |
| LMArena Japanese | 1381 | 1299 |
| LMArena Korean | 1372 | 1304 |
| LMArena Russian | 1430 | 1353 |
| LMArena Spanish | 1431 | 1377 |
Instruction Following MiniMax-M3 leads
MiniMax-M3: 75.7 (#56), Qwen2.5-Max: 71.5 (#147)
| Benchmark | MiniMax-M3 | Qwen2.5-Max |
|---|---|---|
| LMArena Instruction Following | 1433 | 1335 |
| LiveBench Instruction Following | — | 75.3% |
Long Context MiniMax-M3 leads
MiniMax-M3: 44.2 (#68), Qwen2.5-Max: 41.3 (#137)
| Benchmark | MiniMax-M3 | Qwen2.5-Max |
|---|---|---|
| LMArena Longer Query | 1445 | 1358 |
Writing & Preference MiniMax-M3 leads
MiniMax-M3: 62.3 (#79), Qwen2.5-Max: 55.6 (#141)
| Benchmark | MiniMax-M3 | Qwen2.5-Max |
|---|---|---|
| LMArena Text | 1433 | 1367 |
| LMArena Creative Writing | 1405 | 1339 |
| LMArena Multi-Turn | 1441 | 1364 |
| Short-Story Creative Writing | — | 72.9% |
| EQ-Bench 4 | 1150 | — |
| LiveBench Language | — | 56.3% |
Frequently asked questions
Is MiniMax-M3 better than Qwen2.5-Max?
MiniMax-M3 is the stronger model overall, scoring 43.5 to 40.1 on the Noometry Index.
Is MiniMax-M3 or Qwen2.5-Max better for coding?
They score almost the same on coding (41.5 vs 41.2); test both on your own repository before choosing.
How many benchmarks do MiniMax-M3 and Qwen2.5-Max share?
18 benchmarks have published results for both models. MiniMax-M3 has 41 scored results on Noometry and Qwen2.5-Max has 27.