Model comparison
MiniMax-M2 vs Qwen3 32B
Qwen3 32B is the stronger model overall, scoring 39.2 to 37.4 on the Noometry Index. MiniMax-M2 costs 2.3× less per token, which makes it the better buy when Qwen3 32B's lead doesn't matter for your workload.
Last verified . 14 shared benchmarks.
MiniMax
Alibaba (Qwen)Summary
- They share 14 benchmarks with published results for both. MiniMax-M2 scores higher in 3 categories and Qwen3 32B in 6 categories; 6 gaps are clear of the uncertainty.
- The widest gap is in agentic & tool use, where Qwen3 32B leads 32.6 to 25.1.
- MiniMax-M2 is cheaper at $0.30 / $1.20 per million input/output tokens, against $0.70 / $2.80 for Qwen3 32B.
- MiniMax-M2 accepts more context: 205K tokens versus 131K.
Side by side
| MiniMax-M2 | Qwen3 32B | |
|---|---|---|
| Provider | MiniMax | Alibaba (Qwen) |
| Noometry Index | 37.4 | 39.2 |
| Released | 2025-10-27 | 2025-04 |
| Weights | Open | Open |
| Context window | 205K | 131K |
| Max output | 131K | 16K |
| Input $ / M tokens | $0.30 | $0.70 |
| Output $ / M tokens | $1.20 | $2.80 |
| Results tracked | 21 | 26 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding MiniMax-M2 leads
MiniMax-M2: 39.3 (#159), Qwen3 32B: 37.7 (#190)
| Benchmark | MiniMax-M2 | Qwen3 32B |
|---|---|---|
| LMArena Coding | 1370 | 1358 |
| SWE-bench Verified (bash only) | 61% | — |
| Aider Polyglot | — | 40% |
| LMArena WebDev | 1297 | — |
| SciCode | — | 35.4% |
Agentic & Tool Use Qwen3 32B leads
MiniMax-M2: 25.1 (#109), Qwen3 32B: 32.6 (#62)
| Benchmark | MiniMax-M2 | Qwen3 32B |
|---|---|---|
| Terminal-Bench | 30% | — |
| Berkeley Function Calling Leaderboard | — | 48.7% |
| Vending-Bench 2 | 160.6 | — |
Reasoning Too close to call
MiniMax-M2: 19.4 (#258), Qwen3 32B: 20.2 (#241)
| Benchmark | MiniMax-M2 | Qwen3 32B |
|---|---|---|
| Kagi LLM Benchmark | 57.8% | 54.9% |
| LMArena Hard Prompts | 1357 | 1334 |
| NYT Connections (extended) | 14.8% | — |
| CritPt | — | 0.3% |
| Chess Puzzles | — | 5% |
| DTBench | — | 67.5% |
| LMCA | — | 17.3% |
| Epoch Capabilities Index | — | 138.51 |
Math Qwen3 32B leads
MiniMax-M2: 37.3 (#160), Qwen3 32B: 39.7 (#99)
| Benchmark | MiniMax-M2 | Qwen3 32B |
|---|---|---|
| LMArena Math | 1352 | 1399 |
| OTIS Mock AIME 2024-2025 | — | 66.9% |
Knowledge Qwen3 32B leads
MiniMax-M2: 37.0 (#163), Qwen3 32B: 40.0 (#125)
| Benchmark | MiniMax-M2 | Qwen3 32B |
|---|---|---|
| LMArena Expert | 1337 | 1362 |
| GPQA Diamond | — | 65.7% |
| Vectara Hallucination Rate | — | 5.9% |
Multilingual Too close to call
MiniMax-M2: 45.3 (#171), Qwen3 32B: 45.6 (#167)
| Benchmark | MiniMax-M2 | Qwen3 32B |
|---|---|---|
| LMArena Non-English | 1313 | 1317 |
| LMArena Chinese | 1366 | 1357 |
| LMArena German | 1355 | 1341 |
| LMArena Russian | 1331 | 1311 |
| LMArena French | 1335 | — |
| LMArena Spanish | 1326 | — |
Instruction Following MiniMax-M2 leads
MiniMax-M2: 70.2 (#166), Qwen3 32B: 68.9 (#179)
| Benchmark | MiniMax-M2 | Qwen3 32B |
|---|---|---|
| LMArena Instruction Following | 1328 | 1305 |
Long Context Qwen3 32B leads
MiniMax-M2: 40.5 (#153), Qwen3 32B: 43.8 (#87)
| Benchmark | MiniMax-M2 | Qwen3 32B |
|---|---|---|
| LMArena Longer Query | 1331 | 1327 |
| Fiction.LiveBench | — | 74.2% |
Writing & Preference Too close to call
MiniMax-M2: 53.0 (#162), Qwen3 32B: 52.9 (#163)
| Benchmark | MiniMax-M2 | Qwen3 32B |
|---|---|---|
| LMArena Text | 1340 | 1340 |
| LMArena Creative Writing | 1286 | 1297 |
| LMArena Multi-Turn | 1361 | 1331 |
Frequently asked questions
Is MiniMax-M2 better than Qwen3 32B?
Qwen3 32B is the stronger model overall, scoring 39.2 to 37.4 on the Noometry Index. MiniMax-M2 costs 2.3× less per token, which makes it the better buy when Qwen3 32B's lead doesn't matter for your workload.
Which is cheaper, MiniMax-M2 or Qwen3 32B?
MiniMax-M2 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-M2 or Qwen3 32B better for coding?
MiniMax-M2 scores higher on coding benchmarks: 39.3 versus 37.7 in the Noometry coding category.
Which has the bigger context window?
MiniMax-M2 does, with 205K tokens against 131K.
How many benchmarks do MiniMax-M2 and Qwen3 32B share?
14 benchmarks have published results for both models. MiniMax-M2 has 21 scored results on Noometry and Qwen3 32B has 26.