Qwen3 32B vs MiMo-V2.5-Pro
Updated October 21, 2026. 33 benchmarks have a result for both. Coding is ranked on the Vibe Coding Index; every other board is shown as published.
Decision reading
Qwen3 32B and MiMo-V2.5-Pro are level for coding, 40.0 against 40.0 on the Vibe Coding Index, so price and context decide.
Pick MiMo-V2.5-Pro when the context window is larger, 1.05M against 131K tokens, or when what matters is SciCode or Terminal-Bench Hard (Artificial Analysis run), where it leads. Pick Qwen3 32B when output tokens cost 2.2x less.
- Shared results
- 33
- Qwen3 32B only
- 9
- MiMo-V2.5-Pro only
- 37
- Like-for-like categories
- 7 of 9
Shared results can support a head-to-head reading. Results only one model has describe coverage, not superiority.
Which one for your work
Each card is decided by one number both models carry. Where only one has it, the card says so and makes no call.
Coding
Vibe Coding Index, four public coding benchmarks
No call
Level: 40.0 against 40.0, inside 2 points.
One benchmark behind each score
Intelligence
AA Intelligence Index
MiMo-V2.5-Pro
MiMo-V2.5-Pro leads, 26.0 against 7.3.
Same board, both measured
Agentic work
AA Agentic Index
No call
Only MiMo-V2.5-Pro has a AA Agentic Index result, 21.3.
Directional only
Reasoning
GPQA Diamond
No call
Only MiMo-V2.5-Pro has a GPQA Diamond result, 82.6%.
Directional only
Knowledge
MMLU Pro
No call
Only MiMo-V2.5-Pro has a MMLU Pro result, 84.6%.
Directional only
Web design
Design Arena: websites
No call
Only MiMo-V2.5-Pro has a Design Arena: websites result, 1267.
Directional only
Long documents
Listed context window
MiMo-V2.5-Pro
MiMo-V2.5-Pro takes 1.05M tokens against 131K.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
Qwen3 32B
Qwen3 32B $0.0002, MiMo-V2.5-Pro $0.0006 per turn.
List prices
Cost per solved task
One SWE-bench Verified task, solved
No call
Only MiMo-V2.5-Pro has the number, $0.03.
Directional only
At a glance
- Coding index
- Qwen3 32B 40.0 from 1 benchmark, MiMo-V2.5-Pro 40.0 from 1 benchmark.
- Widest category
- Agentic work, 64.3 points apart on average over 1 shared benchmark.
- Widest single gap
- Long context reasoning (AA-LCR): Qwen3 32B 0.0% against MiMo-V2.5-Pro 79.7%, 79.7 points apart on a 0 to 100 scale.
- Output price
- Qwen3 32B $0.28 against MiMo-V2.5-Pro $0.609 per million output tokens, 2.2x apart.
- Context window
- Qwen3 32B takes 131K tokens, MiMo-V2.5-Pro 1.05M.
Benchmark by benchmark
First the score: boards each model leads, per category, where both have a result. Then every board, a category to a fold; bars are each result on a 0 to 100 scale, so compare the two in a row, not one row with another.
| Category | Qwen3 32B | Bars | MiMo-V2.5-Pro | Ties |
|---|---|---|---|---|
| Coding | 0 | 4 | ||
| Intelligence | 0 | 1 | ||
| Agentic work | 0 | 1 | ||
| Reasoning | 0 | 5 | ||
| Math | 0 | 2 | ||
| Knowledge | 0 | 4 | ||
| General | 0 | 16 | ||
| All shared boards | 0 | 33 |
CodingLike-for-likeFour of these boards feed the Vibe Coding Index04
| Benchmark | Qwen3 32B | MiMo-V2.5-Pro | Leader |
|---|---|---|---|
| Aider polyglot | 40.0% | Not measured | Only Qwen3 32B measured |
| LMArena WebDev | Not measured | 1480 | Only MiMo-V2.5-Pro measured |
| AA Coding Index | Not measured | 60.2 | Only MiMo-V2.5-Pro measured |
| LiveCodeBench (Vals AI) | Not measured | 81.3% | Only MiMo-V2.5-Pro measured |
| SWE-bench Verified (Vals AI) | Not measured | 74.0% | Only MiMo-V2.5-Pro measured |
| Terminal-Bench 4 (Vals AI) | Not measured | 0.5% | Only MiMo-V2.5-Pro measured |
| Vibe Code Bench (Vals AI) | Not measured | 34.1% | Only MiMo-V2.5-Pro measured |
| Code migration (Vals AI) | Not measured | 21.6% | Only MiMo-V2.5-Pro measured |
| Terminal-Bench 2.1 (Vals AI) | Not measured | 57.3% | Only MiMo-V2.5-Pro measured |
| LiveCodeBench (Artificial Analysis run) | 54.6% | Not measured | Only Qwen3 32B measured |
| SciCode | 36.0% | 50.6% | MiMo-V2.5-Pro |
| Terminal-Bench Hard (Artificial Analysis run) | 3.0% | 43.2% | MiMo-V2.5-Pro |
| ALE-Bench | Not measured | 899.8 | Only MiMo-V2.5-Pro measured |
| LMArena Text: coding | 1357 | 1501 | MiMo-V2.5-Pro |
| LMArena Text: software and IT | 1366 | 1495 | MiMo-V2.5-Pro |
| LMArena WebDev: React | Not measured | 1469 | Only MiMo-V2.5-Pro measured |
| LMArena WebDev: HTML | Not measured | 1473 | Only MiMo-V2.5-Pro measured |
| Neither model has a result on the other 21 boards here | |||
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index01
| Benchmark | Qwen3 32B | MiMo-V2.5-Pro | Leader |
|---|---|---|---|
| AA Intelligence Index | 7.3 | 26.0 | MiMo-V2.5-Pro |
| Epoch Capabilities Index | 138.4 | Not measured | Only Qwen3 32B measured |
Agentic workLike-for-likePublic boards, shown as published, kept out of the index01
| Benchmark | Qwen3 32B | MiMo-V2.5-Pro | Leader |
|---|---|---|---|
| AA Agentic Index | Not measured | 21.3 | Only MiMo-V2.5-Pro measured |
| Finance Agent (Vals AI) | Not measured | 41.5% | Only MiMo-V2.5-Pro measured |
| Tax Agent Bench (Vals AI) | Not measured | 11.5% | Only MiMo-V2.5-Pro measured |
| Legal Research Bench (Vals AI) | Not measured | 15.9% | Only MiMo-V2.5-Pro measured |
| Excel Modeling Benchmark (Vals AI) | Not measured | 55.2% | Only MiMo-V2.5-Pro measured |
| τ²-Bench Telecom | 29.8% | 94.2% | MiMo-V2.5-Pro |
| Neither model has a result on the other 33 boards here | |||
ReasoningLike-for-likePublic boards, shown as published, kept out of the index05
| Benchmark | Qwen3 32B | MiMo-V2.5-Pro | Leader |
|---|---|---|---|
| GPQA Diamond | Not measured | 82.6% | Only MiMo-V2.5-Pro measured |
| GPQA Diamond (Artificial Analysis run) | 66.8% | 86.6% | MiMo-V2.5-Pro |
| Humanity's Last Exam (Artificial Analysis run) | 7.4% | 35.7% | MiMo-V2.5-Pro |
| GPQA Diamond (Epoch AI run) | 65.7% | Not measured | Only Qwen3 32B measured |
| LMArena Text: hard prompts | 1334 | 1486 | MiMo-V2.5-Pro |
| LMArena Text: expert | 1362 | 1501 | MiMo-V2.5-Pro |
| LMArena Text: hard prompts in English | 1350 | 1494 | MiMo-V2.5-Pro |
| Neither model has a result on the other 14 boards here | |||
MathLike-for-likePublic boards, shown as published, kept out of the index02
| Benchmark | Qwen3 32B | MiMo-V2.5-Pro | Leader |
|---|---|---|---|
| AA Math Index | 73.0 | Not measured | Only Qwen3 32B measured |
| AIME 2025 (Artificial Analysis run) | 73.0% | Not measured | Only Qwen3 32B measured |
| OTIS Mock AIME | 66.9% | Not measured | Only Qwen3 32B measured |
| ProofBench | Not measured | 22.0% | Only MiMo-V2.5-Pro measured |
| LMArena Text: math | 1399 | 1481 | MiMo-V2.5-Pro |
| LMArena Text: mathematical work | 1420 | 1491 | MiMo-V2.5-Pro |
| Neither model has a result on the other 7 boards here | |||
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index04
| Benchmark | Qwen3 32B | MiMo-V2.5-Pro | Leader |
|---|---|---|---|
| MMLU Pro | Not measured | 84.6% | Only MiMo-V2.5-Pro measured |
| LegalBench (Vals AI) | Not measured | 77.1% | Only MiMo-V2.5-Pro measured |
| CorpFin (Vals AI) | Not measured | 61.4% | Only MiMo-V2.5-Pro measured |
| TaxEval (Vals AI) | Not measured | 73.8% | Only MiMo-V2.5-Pro measured |
| MedCode (Vals AI) | Not measured | 32.5% | Only MiMo-V2.5-Pro measured |
| MMLU Pro (Artificial Analysis run) | 79.8% | Not measured | Only Qwen3 32B measured |
| LMArena Text: science | 1372 | 1482 | MiMo-V2.5-Pro |
| LMArena Text: medicine and healthcare | 1368 | 1473 | MiMo-V2.5-Pro |
| LMArena Text: legal and government | 1324 | 1474 | MiMo-V2.5-Pro |
| LMArena Text: business and finance | 1318 | 1466 | MiMo-V2.5-Pro |
| Neither model has a result on the other 7 boards here | |||
Web designDirectional onlyPublic boards, shown as published, kept out of the index8 boards, none shared
| Benchmark | Qwen3 32B | MiMo-V2.5-Pro | Leader |
|---|---|---|---|
| Design Arena: websites | Not measured | 1267 | Only MiMo-V2.5-Pro measured |
| Design Arena: UI components | Not measured | 1248 | Only MiMo-V2.5-Pro measured |
| Design Arena: game dev | Not measured | 1261 | Only MiMo-V2.5-Pro measured |
| Design Arena: data viz | Not measured | 1257 | Only MiMo-V2.5-Pro measured |
| Design Arena: 3D | Not measured | 1244 | Only MiMo-V2.5-Pro measured |
| Design Arena: all code categories | Not measured | 1266 | Only MiMo-V2.5-Pro measured |
| Design Arena: SVG | Not measured | 1181 | Only MiMo-V2.5-Pro measured |
| Design Arena: ASCII art | Not measured | 1157 | Only MiMo-V2.5-Pro measured |
| Neither model has a result on the other 1 board here | |||
GeneralLike-for-likePublic boards, shown as published, kept out of the index016
| Benchmark | Qwen3 32B | MiMo-V2.5-Pro | Leader |
|---|---|---|---|
| MedScribe (Vals AI) | Not measured | 83.7% | Only MiMo-V2.5-Pro measured |
| IFBench | 36.3% | 79.9% | MiMo-V2.5-Pro |
| Long context reasoning (AA-LCR) | 0.0% | 79.7% | MiMo-V2.5-Pro |
| DTBench | 67.5% | 84.5% | MiMo-V2.5-Pro |
| LMCA | 17.3% | 29.5% | MiMo-V2.5-Pro |
| Fiction.LiveBench | 74.2% | Not measured | Only Qwen3 32B measured |
| LMArena Text | 1340 | 1464 | MiMo-V2.5-Pro |
| LMArena Text: instruction following | 1305 | 1475 | MiMo-V2.5-Pro |
| LMArena Text: creative writing | 1297 | 1438 | MiMo-V2.5-Pro |
| LMArena Text: English | 1359 | 1477 | MiMo-V2.5-Pro |
| LMArena Text: non-English | 1317 | 1448 | MiMo-V2.5-Pro |
| LMArena Text: Chinese | 1357 | 1507 | MiMo-V2.5-Pro |
| LMArena Text: German | 1341 | 1459 | MiMo-V2.5-Pro |
| LMArena Text: Japanese | Not measured | 1410 | Only MiMo-V2.5-Pro measured |
| LMArena Text: Korean | Not measured | 1435 | Only MiMo-V2.5-Pro measured |
| LMArena Text: French | Not measured | 1487 | Only MiMo-V2.5-Pro measured |
| LMArena Text: Spanish | Not measured | 1469 | Only MiMo-V2.5-Pro measured |
| LMArena Text: Russian | 1311 | 1449 | MiMo-V2.5-Pro |
| LMArena Text: Polish | Not measured | 1464 | Only MiMo-V2.5-Pro measured |
| LMArena Text: multi-turn | 1331 | 1476 | MiMo-V2.5-Pro |
| LMArena Text: longer queries | 1327 | 1481 | MiMo-V2.5-Pro |
| LMArena Text: writing and language | 1298 | 1451 | MiMo-V2.5-Pro |
| LMArena Text: entertainment and media | 1279 | 1438 | MiMo-V2.5-Pro |
| Neither model has a result on the other 23 boards here | |||
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | Qwen3 32B | MiMo-V2.5-Pro | Leader |
|---|---|---|---|
| Neither model has a result on 5 boards here | |||
What each workload costs
Three fixed token mixes at each model's listed price, so per-token rates become a bill.
Chat turn
1K input + 500 output tokens
Qwen3 32B costs less
Repository review
50K input + 3K output tokens
Qwen3 32B costs less
Agent loop
200K input + 10K output tokens
Qwen3 32B costs less
Listed OpenRouter rates, without cache or batch discounts. A workload that does not fit a model's context window would need splitting.
Specification differences
The leading value in a row is marked; a row with a missing value names no leader. Our two rows are marked in red.
| Specification | Qwen3 32B | Xiaomi MiMo-V2.5-Pro |
|---|---|---|
| Maker | Qwen | Xiaomi |
| Weights | Open weights | Open weights |
| Released | April 28, 2025 | April 22, 2026 |
| Context window | 131K | 1.05M |
| Input $/Mtok | $0.08 | $0.3045 |
| Output $/Mtok | $0.28 | $0.609 |
| Coding index | 40.0 | 40.0 |
| Independent measurements behind it | 1 | 1 |
| Cost per solved task | Not computable | $0.03 |
| Source | OpenRouter listing | OpenRouter listing |
Questions
Is Qwen3 32B better than MiMo-V2.5-Pro for coding?
Qwen3 32B and MiMo-V2.5-Pro are level for coding, 40.0 against 40.0 on the Vibe Coding Index, so price and context decide. Pick MiMo-V2.5-Pro when the context window is larger, 1.05M against 131K tokens, or when what matters is SciCode or Terminal-Bench Hard (Artificial Analysis run), where it leads. Pick Qwen3 32B when output tokens cost 2.2x less.
Which is cheaper, Qwen3 32B or MiMo-V2.5-Pro?
Qwen3 32B costs $0.08 per million input tokens and $0.28 per million output tokens; MiMo-V2.5-Pro costs $0.3045 and $0.609. Cost per solved task is known for MiMo-V2.5-Pro only.
Which has the bigger context window, Qwen3 32B or MiMo-V2.5-Pro?
MiMo-V2.5-Pro, with 1.05M tokens against 131K.
How many benchmarks is the Qwen3 32B vs MiMo-V2.5-Pro comparison based on?
33 benchmarks have a result for both models as of October 21, 2026. Qwen3 32B's coding index rests on 1 independent benchmark and MiMo-V2.5-Pro's on 1. Where only one model has a result, the page says so instead of naming a winner.