Model comparison
MiniMax-M3 vs Qwen Max
MiniMax-M3 is the stronger model overall, scoring 43.8 to 34.7 on the Noometry Index.
Last verified . 19 shared benchmarks.
MiniMax
Alibaba (Qwen)Summary
- They share 19 benchmarks with published results for both. MiniMax-M3 scores higher in 8 categories and Qwen Max in 0 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where MiniMax-M3 leads 58.4 to 30.3.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 71.1% for MiniMax-M3 and 16.1% for Qwen Max.
- MiniMax-M3 is cheaper at $0.30 / $1.20 per million input/output tokens, against $1.60 / $6.40 for Qwen Max.
- MiniMax-M3 accepts more context: 1M tokens versus 33K.
- MiniMax-M3 has downloadable open weights; the other is API-only.
Side by side
| MiniMax-M3 | Qwen Max | |
|---|---|---|
| Provider | MiniMax | Alibaba (Qwen) |
| Noometry Index | 43.8 | 34.7 |
| Released | 2026-06-01 | 2024-04-03 |
| Weights | Open | Proprietary |
| Context window | 1M | 33K |
| Max output | 512K | 8K |
| Input $ / M tokens | $0.30 | $1.60 |
| Output $ / M tokens | $1.20 | $6.40 |
| Results tracked | 41 | 23 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding MiniMax-M3 leads
MiniMax-M3: 41.8 (#118), Qwen Max: 30.7 (#292)
| Benchmark | MiniMax-M3 | Qwen Max |
|---|---|---|
| LMArena Coding | 1469 | 1288 |
| FrontierCode | 14.7% | — |
| Aider Polyglot | — | 21.8% |
| LMArena WebDev | 1482 | — |
| SciCode | 47.1% | — |
| ALE-Bench | 640.02 | — |
Agentic & Tool Use Not comparable
MiniMax-M3: 22.6 (#130), Qwen Max: —
| Benchmark | MiniMax-M3 | Qwen Max |
|---|---|---|
| APEX-Agents | 37.7% | — |
| OSWorld 2.0 | 4.6% | — |
| GBAEval | 0.9% | — |
| Vending-Bench 2 | 2,158 | — |
Reasoning MiniMax-M3 leads
MiniMax-M3: 30.1 (#87), Qwen Max: 25.1 (#151)
| Benchmark | MiniMax-M3 | Qwen Max |
|---|---|---|
| LMArena Hard Prompts | 1447 | 1269 |
| SimpleBench | 45.8% | — |
| NYT Connections (extended) | 65.1% | — |
| CritPt | 3.7% | — |
| Chess Puzzles | 14% | — |
| Mystery Game Puzzles | 8% | — |
| DTBench | 78.9% | — |
| LMCA | 33.7% | — |
| Surface Evolver Bench | 55% | — |
| Epoch Capabilities Index | 146.95 | — |
| ForecastBench | 61.4 | — |
Math MiniMax-M3 leads
MiniMax-M3: 40.0 (#95), Qwen Max: 22.3 (#276)
| Benchmark | MiniMax-M3 | Qwen Max |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 71.1% | 16.1% |
| LMArena Math | 1429 | 1275 |
| ProofBench | 18% | — |
| MATH Level 5 | — | 67.2% |
| FrontierMath (Feb 2025 set) | — | 1% |
Knowledge MiniMax-M3 leads
MiniMax-M3: 58.4 (#35), Qwen Max: 30.3 (#228)
| Benchmark | MiniMax-M3 | Qwen Max |
|---|---|---|
| GPQA Diamond | 90.9% | 56.1% |
| LMArena Expert | 1461 | 1248 |
Multimodal Not comparable
MiniMax-M3: 40.2 (#51), Qwen Max: —
| Benchmark | MiniMax-M3 | Qwen Max |
|---|---|---|
| LMArena Vision | 1253 | — |
| LMArena Document | 1435 | — |
Multilingual MiniMax-M3 leads
MiniMax-M3: 53.0 (#75), Qwen Max: 41.8 (#202)
| Benchmark | MiniMax-M3 | Qwen Max |
|---|---|---|
| LMArena Non-English | 1420 | 1263 |
| LMArena Chinese | 1463 | 1254 |
| LMArena French | 1447 | 1330 |
| LMArena German | 1426 | 1254 |
| LMArena Japanese | 1381 | 1205 |
| LMArena Korean | 1372 | 1142 |
| LMArena Russian | 1428 | 1274 |
| LMArena Spanish | 1432 | 1290 |
Instruction Following MiniMax-M3 leads
MiniMax-M3: 75.5 (#62), Qwen Max: 66.5 (#208)
| Benchmark | MiniMax-M3 | Qwen Max |
|---|---|---|
| LMArena Instruction Following | 1433 | 1262 |
Long Context MiniMax-M3 leads
MiniMax-M3: 44.2 (#72), Qwen Max: 39.4 (#180)
| Benchmark | MiniMax-M3 | Qwen Max |
|---|---|---|
| LMArena Longer Query | 1445 | 1288 |
| Fiction.LiveBench | — | 66.7% |
Writing & Preference MiniMax-M3 leads
MiniMax-M3: 62.1 (#83), Qwen Max: 47.8 (#205)
| Benchmark | MiniMax-M3 | Qwen Max |
|---|---|---|
| LMArena Text | 1433 | 1282 |
| LMArena Creative Writing | 1404 | 1248 |
| LMArena Multi-Turn | 1442 | 1277 |
| EQ-Bench 4 | 1150 | — |
Frequently asked questions
Is MiniMax-M3 better than Qwen Max?
MiniMax-M3 is the stronger model overall, scoring 43.8 to 34.7 on the Noometry Index.
Which is cheaper, MiniMax-M3 or Qwen Max?
MiniMax-M3 is cheaper. It lists at $0.30 per million input tokens and $1.20 per million output tokens; Qwen Max lists at $1.60 and $6.40.
Is MiniMax-M3 or Qwen Max better for coding?
MiniMax-M3 scores higher on coding benchmarks: 41.8 versus 30.7 in the Noometry coding category.
Which has the bigger context window?
MiniMax-M3 does, with 1M tokens against 33K.
How many benchmarks do MiniMax-M3 and Qwen Max share?
19 benchmarks have published results for both models. MiniMax-M3 has 41 scored results on Noometry and Qwen Max has 23.