Model comparison
MiniMax-M3 vs Qwen3 32B
MiniMax-M3 is the stronger model overall, scoring 43.8 to 39.2 on the Noometry Index.
Last verified . 21 shared benchmarks.
MiniMax
Alibaba (Qwen)Summary
- They share 21 benchmarks with published results for both. MiniMax-M3 scores higher in 8 categories and Qwen3 32B in 1 category; 7 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where MiniMax-M3 leads 58.4 to 40.0.
- The biggest single-benchmark swing is GPQA Diamond: 90.9% for MiniMax-M3 and 65.7% for Qwen3 32B.
- MiniMax-M3 is cheaper at $0.30 / $1.20 per million input/output tokens, against $0.70 / $2.80 for Qwen3 32B.
- MiniMax-M3 accepts more context: 1M tokens versus 131K.
Side by side
| MiniMax-M3 | Qwen3 32B | |
|---|---|---|
| Provider | MiniMax | Alibaba (Qwen) |
| Noometry Index | 43.8 | 39.2 |
| Released | 2026-06-01 | 2025-04 |
| Weights | Open | Open |
| Context window | 1M | 131K |
| Max output | 512K | 16K |
| Input $ / M tokens | $0.30 | $0.70 |
| Output $ / M tokens | $1.20 | $2.80 |
| Results tracked | 41 | 26 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding MiniMax-M3 leads
MiniMax-M3: 41.8 (#118), Qwen3 32B: 37.7 (#190)
| Benchmark | MiniMax-M3 | Qwen3 32B |
|---|---|---|
| SciCode | 47.1% | 35.4% |
| LMArena Coding | 1469 | 1358 |
| FrontierCode | 14.7% | — |
| Aider Polyglot | — | 40% |
| LMArena WebDev | 1482 | — |
| ALE-Bench | 640.02 | — |
Agentic & Tool Use Qwen3 32B leads
MiniMax-M3: 22.6 (#130), Qwen3 32B: 32.6 (#62)
| Benchmark | MiniMax-M3 | Qwen3 32B |
|---|---|---|
| APEX-Agents | 37.7% | — |
| Berkeley Function Calling Leaderboard | — | 48.7% |
| OSWorld 2.0 | 4.6% | — |
| GBAEval | 0.9% | — |
| Vending-Bench 2 | 2,158 | — |
Reasoning MiniMax-M3 leads
MiniMax-M3: 30.1 (#87), Qwen3 32B: 20.2 (#241)
| Benchmark | MiniMax-M3 | Qwen3 32B |
|---|---|---|
| CritPt | 3.7% | 0.3% |
| Chess Puzzles | 14% | 5% |
| LMArena Hard Prompts | 1447 | 1334 |
| DTBench | 78.9% | 67.5% |
| LMCA | 33.7% | 17.3% |
| Epoch Capabilities Index | 146.95 | 138.51 |
| SimpleBench | 45.8% | — |
| Kagi LLM Benchmark | — | 54.9% |
| NYT Connections (extended) | 65.1% | — |
| Mystery Game Puzzles | 8% | — |
| Surface Evolver Bench | 55% | — |
| ForecastBench | 61.4 | — |
Math Too close to call
MiniMax-M3: 40.0 (#95), Qwen3 32B: 39.7 (#99)
| Benchmark | MiniMax-M3 | Qwen3 32B |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 71.1% | 66.9% |
| LMArena Math | 1429 | 1399 |
| ProofBench | 18% | — |
Knowledge MiniMax-M3 leads
MiniMax-M3: 58.4 (#35), Qwen3 32B: 40.0 (#125)
| Benchmark | MiniMax-M3 | Qwen3 32B |
|---|---|---|
| GPQA Diamond | 90.9% | 65.7% |
| LMArena Expert | 1461 | 1362 |
| Vectara Hallucination Rate | — | 5.9% |
Multimodal Not comparable
MiniMax-M3: 40.2 (#51), Qwen3 32B: —
| Benchmark | MiniMax-M3 | Qwen3 32B |
|---|---|---|
| LMArena Vision | 1253 | — |
| LMArena Document | 1435 | — |
Multilingual MiniMax-M3 leads
MiniMax-M3: 53.0 (#75), Qwen3 32B: 45.6 (#167)
| Benchmark | MiniMax-M3 | Qwen3 32B |
|---|---|---|
| LMArena Non-English | 1420 | 1317 |
| LMArena Chinese | 1463 | 1357 |
| LMArena German | 1426 | 1341 |
| LMArena Russian | 1428 | 1311 |
| LMArena French | 1447 | — |
| LMArena Japanese | 1381 | — |
| LMArena Korean | 1372 | — |
| LMArena Spanish | 1432 | — |
Instruction Following MiniMax-M3 leads
MiniMax-M3: 75.5 (#62), Qwen3 32B: 68.9 (#179)
| Benchmark | MiniMax-M3 | Qwen3 32B |
|---|---|---|
| LMArena Instruction Following | 1433 | 1305 |
Long Context Too close to call
MiniMax-M3: 44.2 (#72), Qwen3 32B: 43.8 (#87)
| Benchmark | MiniMax-M3 | Qwen3 32B |
|---|---|---|
| LMArena Longer Query | 1445 | 1327 |
| Fiction.LiveBench | — | 74.2% |
Writing & Preference MiniMax-M3 leads
MiniMax-M3: 62.1 (#83), Qwen3 32B: 52.9 (#163)
| Benchmark | MiniMax-M3 | Qwen3 32B |
|---|---|---|
| LMArena Text | 1433 | 1340 |
| LMArena Creative Writing | 1404 | 1297 |
| LMArena Multi-Turn | 1442 | 1331 |
| EQ-Bench 4 | 1150 | — |
Frequently asked questions
Is MiniMax-M3 better than Qwen3 32B?
MiniMax-M3 is the stronger model overall, scoring 43.8 to 39.2 on the Noometry Index.
Which is cheaper, MiniMax-M3 or Qwen3 32B?
MiniMax-M3 is cheaper. It lists at $0.30 per million input tokens and $1.20 per million output tokens; Qwen3 32B lists at $0.70 and $2.80.
Is MiniMax-M3 or Qwen3 32B better for coding?
MiniMax-M3 scores higher on coding benchmarks: 41.8 versus 37.7 in the Noometry coding category.
Which has the bigger context window?
MiniMax-M3 does, with 1M tokens against 131K.
How many benchmarks do MiniMax-M3 and Qwen3 32B share?
21 benchmarks have published results for both models. MiniMax-M3 has 41 scored results on Noometry and Qwen3 32B has 26.