Model comparison
MiMo-V2.5 vs Qwen3 32B
MiMo-V2.5 is the stronger model overall, scoring 43.4 to 39.2 on the Noometry Index.
Last verified . 15 shared benchmarks.
Alibaba (Qwen)Summary
- They share 15 benchmarks with published results for both. MiMo-V2.5 scores higher in 7 categories and Qwen3 32B in 1 category; 6 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where MiMo-V2.5 leads 61.6 to 52.9.
- The biggest single-benchmark swing is SciCode: 43.1% for MiMo-V2.5 and 35.4% for Qwen3 32B.
- MiMo-V2.5 is cheaper at $0.14 / $0.28 per million input/output tokens, against $0.70 / $2.80 for Qwen3 32B.
- MiMo-V2.5 accepts more context: 1.05M tokens versus 131K.
Side by side
| MiMo-V2.5 | Qwen3 32B | |
|---|---|---|
| Provider | Xiaomi | Alibaba (Qwen) |
| Noometry Index | 43.4 | 39.2 |
| Released | 2026-04-22 | 2025-04 |
| Weights | Open | Open |
| Context window | 1.05M | 131K |
| Max output | 131K | 16K |
| Input $ / M tokens | $0.14 | $0.70 |
| Output $ / M tokens | $0.28 | $2.80 |
| Results tracked | 23 | 26 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding MiMo-V2.5 leads
MiMo-V2.5: 43.9 (#81), Qwen3 32B: 37.7 (#190)
| Benchmark | MiMo-V2.5 | Qwen3 32B |
|---|---|---|
| SciCode | 43.1% | 35.4% |
| LMArena Coding | 1469 | 1358 |
| Aider Polyglot | — | 40% |
| LMArena WebDev | 1438 | — |
| ALE-Bench | 513.95 | — |
Agentic & Tool Use Not comparable
MiMo-V2.5: —, Qwen3 32B: 32.6 (#62)
| Benchmark | MiMo-V2.5 | Qwen3 32B |
|---|---|---|
| Berkeley Function Calling Leaderboard | — | 48.7% |
Reasoning MiMo-V2.5 leads
MiMo-V2.5: 28.6 (#101), Qwen3 32B: 20.2 (#241)
| Benchmark | MiMo-V2.5 | Qwen3 32B |
|---|---|---|
| CritPt | 3.7% | 0.3% |
| LMArena Hard Prompts | 1450 | 1334 |
| Kagi LLM Benchmark | — | 54.9% |
| Chess Puzzles | — | 5% |
| DTBench | — | 67.5% |
| LMCA | — | 17.3% |
| Epoch Capabilities Index | — | 138.51 |
Math Qwen3 32B leads
MiMo-V2.5: 36.8 (#163), Qwen3 32B: 39.7 (#99)
| Benchmark | MiMo-V2.5 | Qwen3 32B |
|---|---|---|
| LMArena Math | 1436 | 1399 |
| OTIS Mock AIME 2024-2025 | — | 66.9% |
| ProofBench | 16% | — |
Knowledge Too close to call
MiMo-V2.5: 40.8 (#115), Qwen3 32B: 40.0 (#125)
| Benchmark | MiMo-V2.5 | Qwen3 32B |
|---|---|---|
| LMArena Expert | 1460 | 1362 |
| GPQA Diamond | — | 65.7% |
| Vectara Hallucination Rate | — | 5.9% |
Multimodal Not comparable
MiMo-V2.5: 39.8 (#54), Qwen3 32B: —
| Benchmark | MiMo-V2.5 | Qwen3 32B |
|---|---|---|
| LMArena Vision | 1247 | — |
Multilingual MiMo-V2.5 leads
MiMo-V2.5: 51.9 (#99), Qwen3 32B: 45.6 (#167)
| Benchmark | MiMo-V2.5 | Qwen3 32B |
|---|---|---|
| LMArena Non-English | 1404 | 1317 |
| LMArena Chinese | 1468 | 1357 |
| LMArena German | 1421 | 1341 |
| LMArena Russian | 1395 | 1311 |
| LMArena French | 1447 | — |
| LMArena Japanese | 1306 | — |
| LMArena Korean | 1363 | — |
| LMArena Spanish | 1416 | — |
Instruction Following MiMo-V2.5 leads
MiMo-V2.5: 75.5 (#60), Qwen3 32B: 68.9 (#179)
| Benchmark | MiMo-V2.5 | Qwen3 32B |
|---|---|---|
| LMArena Instruction Following | 1434 | 1305 |
Long Context Too close to call
MiMo-V2.5: 44.2 (#73), Qwen3 32B: 43.8 (#87)
| Benchmark | MiMo-V2.5 | Qwen3 32B |
|---|---|---|
| LMArena Longer Query | 1445 | 1327 |
| Fiction.LiveBench | — | 74.2% |
Writing & Preference MiMo-V2.5 leads
MiMo-V2.5: 61.6 (#86), Qwen3 32B: 52.9 (#163)
| Benchmark | MiMo-V2.5 | Qwen3 32B |
|---|---|---|
| LMArena Text | 1428 | 1340 |
| LMArena Creative Writing | 1393 | 1297 |
| LMArena Multi-Turn | 1445 | 1331 |
Frequently asked questions
Is MiMo-V2.5 better than Qwen3 32B?
MiMo-V2.5 is the stronger model overall, scoring 43.4 to 39.2 on the Noometry Index.
Which is cheaper, MiMo-V2.5 or Qwen3 32B?
MiMo-V2.5 is cheaper. It lists at $0.14 per million input tokens and $0.28 per million output tokens; Qwen3 32B lists at $0.70 and $2.80.
Is MiMo-V2.5 or Qwen3 32B better for coding?
MiMo-V2.5 scores higher on coding benchmarks: 43.9 versus 37.7 in the Noometry coding category.
Which has the bigger context window?
MiMo-V2.5 does, with 1.05M tokens against 131K.
How many benchmarks do MiMo-V2.5 and Qwen3 32B share?
15 benchmarks have published results for both models. MiMo-V2.5 has 23 scored results on Noometry and Qwen3 32B has 26.