# MiMo-V2.5-Pro vs Qwen3 32B

> MiMo-V2.5-Pro is the stronger model overall, scoring 44.8 to 38.8 on the Noometry Index.

- Canonical page: https://noometry.com/compare/mimo-v2-5-pro-vs-qwen3-32b
- Last updated: 2026-10-10
- Shared benchmarks: 17

## Summary

- They share 17 benchmarks with published results for both. MiMo-V2.5-Pro scores higher in 8 categories and Qwen3 32B in 0 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where MiMo-V2.5-Pro leads 65.5 to 53.1.
- The biggest single-benchmark swing is DTBench: 84.5% for MiMo-V2.5-Pro and 67.5% for Qwen3 32B.
- MiMo-V2.5-Pro is cheaper at $0.43 / $0.87 per million input/output tokens, against $0.70 / $2.80 for Qwen3 32B.
- MiMo-V2.5-Pro accepts more context: 1.05M tokens versus 131K.

## Snapshot

| | MiMo-V2.5-Pro | Qwen3 32B |
|---|---|---|
| Provider | Xiaomi | Alibaba (Qwen) |
| Noometry Index | 44.8 | 38.8 |
| Rank | 70 | 166 |
| Context | 1.05M | 131K |
| Input $/M | $0.43 | $0.70 |
| Output $/M | $0.87 | $2.80 |
| Weights | Open | Open |

## Coding

- MiMo-V2.5-Pro: 47.0 (#58)
- Qwen3 32B: 37.3 (#184)

| Benchmark | MiMo-V2.5-Pro | Qwen3 32B |
|---|---|---|
| SciCode | 50.2% | 35.4% |
| LMArena Coding | 1502 | 1358 |
| Aider Polyglot | — | 40% |
| LMArena WebDev | 1477 | — |
| ALE-Bench | 899.8 | — |

## Agentic & Tool Use

- MiMo-V2.5-Pro: —
- Qwen3 32B: 32.3 (#58)

| Benchmark | MiMo-V2.5-Pro | Qwen3 32B |
|---|---|---|
| Berkeley Function Calling Leaderboard | — | 48.7% |

## Reasoning

- MiMo-V2.5-Pro: 25.9 (#124)
- Qwen3 32B: 19.6 (#233)

| Benchmark | MiMo-V2.5-Pro | Qwen3 32B |
|---|---|---|
| CritPt | 4% | 0.3% |
| LMArena Hard Prompts | 1487 | 1335 |
| DTBench | 84.5% | 67.5% |
| LMCA | 29.5% | 17.3% |
| Kagi LLM Benchmark | — | 54.9% |
| NYT Connections (extended) | 34.4% | — |
| Chess Puzzles | — | 5% |
| Epoch Capabilities Index | — | 138.51 |

## Math

- MiMo-V2.5-Pro: 38.6 (#84)
- Qwen3 32B: 37.9 (#89)

| Benchmark | MiMo-V2.5-Pro | Qwen3 32B |
|---|---|---|
| LMArena Math | 1480 | 1399 |
| OTIS Mock AIME 2024-2025 | — | 66.9% |
| ProofBench | 22% | — |

## Knowledge

- MiMo-V2.5-Pro: 41.5 (#100)
- Qwen3 32B: 40.0 (#112)

| Benchmark | MiMo-V2.5-Pro | Qwen3 32B |
|---|---|---|
| LMArena Expert | 1501 | 1362 |
| GPQA Diamond | — | 65.7% |
| Vectara Hallucination Rate | — | 5.9% |

## Multilingual

- MiMo-V2.5-Pro: 55.3 (#33)
- Qwen3 32B: 45.8 (#162)

| Benchmark | MiMo-V2.5-Pro | Qwen3 32B |
|---|---|---|
| LMArena Non-English | 1449 | 1317 |
| LMArena Chinese | 1507 | 1357 |
| LMArena German | 1455 | 1341 |
| LMArena Russian | 1450 | 1311 |
| LMArena French | 1489 | — |
| LMArena Japanese | 1418 | — |
| LMArena Korean | 1436 | — |
| LMArena Spanish | 1470 | — |

## Instruction Following

- MiMo-V2.5-Pro: 77.6 (#21)
- Qwen3 32B: 69.1 (#174)

| Benchmark | MiMo-V2.5-Pro | Qwen3 32B |
|---|---|---|
| LMArena Instruction Following | 1477 | 1305 |

## Long Context

- MiMo-V2.5-Pro: 45.4 (#38)
- Qwen3 32B: 43.7 (#83)

| Benchmark | MiMo-V2.5-Pro | Qwen3 32B |
|---|---|---|
| LMArena Longer Query | 1482 | 1327 |
| Fiction.LiveBench | — | 74.2% |

## Writing & Preference

- MiMo-V2.5-Pro: 65.5 (#47)
- Qwen3 32B: 53.1 (#158)

| Benchmark | MiMo-V2.5-Pro | Qwen3 32B |
|---|---|---|
| LMArena Text | 1465 | 1340 |
| LMArena Creative Writing | 1440 | 1297 |
| LMArena Multi-Turn | 1477 | 1331 |
| EQ-Bench Creative Writing | 1493 | — |
| EQ-Bench 4 | 1208 | — |

## FAQ

### Is MiMo-V2.5-Pro better than Qwen3 32B?

MiMo-V2.5-Pro is the stronger model overall, scoring 44.8 to 38.8 on the Noometry Index.

### Which is cheaper, MiMo-V2.5-Pro or Qwen3 32B?

MiMo-V2.5-Pro is cheaper. It lists at $0.43 per million input tokens and $0.87 per million output tokens; Qwen3 32B lists at $0.70 and $2.80.

### Is MiMo-V2.5-Pro or Qwen3 32B better for coding?

MiMo-V2.5-Pro scores higher on coding benchmarks: 47.0 versus 37.3 in the Noometry coding category.

### Which has the bigger context window?

MiMo-V2.5-Pro does, with 1.05M tokens against 131K.

### How many benchmarks do MiMo-V2.5-Pro and Qwen3 32B share?

17 benchmarks have published results for both models. MiMo-V2.5-Pro has 27 scored results on Noometry and Qwen3 32B has 26.