Gemma 4 31B vs MiMo-V2.5-Pro
Gemma 4 31B leads MiMo-V2.5-Pro on coding, 39.0 to 31.6 on the Vibe Coding Index, and Gemma 4 31B costs 2.3x less per output token ($0.40 vs $0.9152 per million). Updated October 22, 2026; 41 benchmarks have a result for both.
Gemma 4 31B or MiMo-V2.5-Pro: which to pick
Pick Gemma 4 31B for coding: it scores 39.0 on the Vibe Coding Index against 31.6 for MiMo-V2.5-Pro, though on 2 benchmarks against 4.
Pick MiMo-V2.5-Pro for a larger context window (1.05M against 262K tokens) and work measured by LMArena WebDev or AA Coding Index, where it leads.
- Shared results
- 41
- Gemma 4 31B only
- 30
- MiMo-V2.5-Pro only
- 29
- Like-for-like categories
- 7 of 9
Only benchmarks that tested both models count toward a pick; a result only one model has is shown but not compared.
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, six public coding benchmarks
Gemma 4 31B
Gemma 4 31B scores 39.0 against 31.6.
2 and 4 benchmarks behind the scores
Intelligence
AA Intelligence Index
MiMo-V2.5-Pro
MiMo-V2.5-Pro leads, 26.0 against 14.7.
Same board, both measured
Agentic work
AA Agentic Index
MiMo-V2.5-Pro
MiMo-V2.5-Pro leads, 21.3 against 4.2.
Same board, both measured
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, 1266.
Directional only
Long documents
Listed context window
MiMo-V2.5-Pro
MiMo-V2.5-Pro takes 1.05M tokens against 262K.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
Gemma 4 31B
Gemma 4 31B $0.0003, MiMo-V2.5-Pro $0.0009 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
- Gemma 4 31B 39.0 from 2 benchmarks, MiMo-V2.5-Pro 31.6 from 4 benchmarks.
- Widest category
- Agentic work, 22.9 points apart on average over 2 shared benchmarks.
- Widest single gap
- τ²-Bench Telecom: Gemma 4 31B 65.5% against MiMo-V2.5-Pro 94.2%, 28.7 points apart on a 0 to 100 scale.
- Output price
- Gemma 4 31B $0.40 against MiMo-V2.5-Pro $0.9152 per million output tokens, 2.3x apart.
- Context window
- Gemma 4 31B takes 262K 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 | Gemma 4 31B | Bars | MiMo-V2.5-Pro | Ties |
|---|---|---|---|---|
| Coding | 1 | 8 | ||
| Intelligence | 0 | 1 | ||
| Agentic work | 0 | 2 | ||
| Reasoning | 0 | 5 | ||
| Math | 0 | 2 | ||
| Knowledge | 0 | 4 | ||
| General | 2 | 16 | ||
| All shared boards | 3 | 38 |
CodingLike-for-likeSix of these boards feed the Vibe Coding Index18
| Benchmark | Gemma 4 31B | MiMo-V2.5-Pro | Leader |
|---|---|---|---|
| LiveBench coding | 49.0% | Not measured | Only Gemma 4 31B measured |
| LMArena WebDev | 1367 | 1480 | MiMo-V2.5-Pro |
| AA Coding Index | 43.4 | 60.2 | MiMo-V2.5-Pro |
| 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 |
| SciCode | 45.5% | 50.6% | MiMo-V2.5-Pro |
| Terminal-Bench Hard (Artificial Analysis run) | 36.4% | 43.2% | MiMo-V2.5-Pro |
| WeirdML | 52.3% | Not measured | Only Gemma 4 31B measured |
| ALE-Bench | 925.5 | 899.8 | Gemma 4 31B |
| LMArena Text: coding | 1459 | 1501 | MiMo-V2.5-Pro |
| LMArena Text: software and IT | 1463 | 1495 | MiMo-V2.5-Pro |
| LMArena WebDev: React | 1359 | 1469 | MiMo-V2.5-Pro |
| LMArena WebDev: HTML | 1380 | 1473 | MiMo-V2.5-Pro |
| SciCode | 8.9% | Not measured | Only Gemma 4 31B 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 | Gemma 4 31B | MiMo-V2.5-Pro | Leader |
|---|---|---|---|
| AA Intelligence Index | 14.7 | 26.0 | MiMo-V2.5-Pro |
| Epoch Capabilities Index | 143.0 | Not measured | Only Gemma 4 31B measured |
Agentic workLike-for-likePublic boards, shown as published, kept out of the index02
| Benchmark | Gemma 4 31B | MiMo-V2.5-Pro | Leader |
|---|---|---|---|
| AA Agentic Index | 4.2 | 21.3 | MiMo-V2.5-Pro |
| LiveBench research | 44.4% | Not measured | Only Gemma 4 31B measured |
| LiveBench browser | 4.0% | Not measured | Only Gemma 4 31B measured |
| LiveBench long agentic | 13.3% | Not measured | Only Gemma 4 31B 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 | 65.5% | 94.2% | MiMo-V2.5-Pro |
| Neither model has a result on the other 32 boards here | |||
ReasoningLike-for-likePublic boards, shown as published, kept out of the index05
| Benchmark | Gemma 4 31B | MiMo-V2.5-Pro | Leader |
|---|---|---|---|
| GPQA Diamond | Not measured | 82.6% | Only MiMo-V2.5-Pro measured |
| LiveBench reasoning | 68.3% | Not measured | Only Gemma 4 31B measured |
| GPQA Diamond (Artificial Analysis run) | 85.7% | 86.6% | MiMo-V2.5-Pro |
| Humanity's Last Exam (Artificial Analysis run) | 23.6% | 35.7% | MiMo-V2.5-Pro |
| GPQA Diamond (Epoch AI run) | 75.8% | Not measured | Only Gemma 4 31B measured |
| LMArena Text: hard prompts | 1448 | 1486 | MiMo-V2.5-Pro |
| LMArena Text: expert | 1464 | 1501 | MiMo-V2.5-Pro |
| LMArena Text: hard prompts in English | 1456 | 1494 | MiMo-V2.5-Pro |
| LMArena Vision: diagrams | 1283 | Not measured | Only Gemma 4 31B measured |
| LMArena Vision: homework | 1310 | Not measured | Only Gemma 4 31B measured |
| Kagi LLM Benchmark | 63.5% | Not measured | Only Gemma 4 31B measured |
| Neither model has a result on the other 11 boards here | |||
MathLike-for-likePublic boards, shown as published, kept out of the index02
| Benchmark | Gemma 4 31B | MiMo-V2.5-Pro | Leader |
|---|---|---|---|
| OTIS Mock AIME | 73.3% | Not measured | Only Gemma 4 31B measured |
| ProofBench | Not measured | 22.0% | Only MiMo-V2.5-Pro measured |
| LMArena Text: math | 1463 | 1481 | MiMo-V2.5-Pro |
| LMArena Text: mathematical work | 1465 | 1491 | MiMo-V2.5-Pro |
| Neither model has a result on the other 9 boards here | |||
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index04
| Benchmark | Gemma 4 31B | 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 |
| CaseLaw (Vals AI) | 52.6% | Not measured | Only Gemma 4 31B 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 |
| MortgageTax (Vals AI) | 61.4% | Not measured | Only Gemma 4 31B measured |
| MedCode (Vals AI) | Not measured | 32.5% | Only MiMo-V2.5-Pro measured |
| SimpleQA Verified | 10.4% | Not measured | Only Gemma 4 31B measured |
| LMArena Text: science | 1457 | 1482 | MiMo-V2.5-Pro |
| LMArena Text: medicine and healthcare | 1451 | 1473 | MiMo-V2.5-Pro |
| LMArena Text: legal and government | 1438 | 1474 | MiMo-V2.5-Pro |
| LMArena Text: business and finance | 1430 | 1466 | MiMo-V2.5-Pro |
| FACTS Benchmark Suite | 45.3% | Not measured | Only Gemma 4 31B measured |
| Neither model has a result on the other 5 boards here | |||
Web designDirectional onlyPublic boards, shown as published, kept out of the index9 boards, none shared
| Benchmark | Gemma 4 31B | MiMo-V2.5-Pro | Leader |
|---|---|---|---|
| Design Arena: websites | Not measured | 1266 | Only MiMo-V2.5-Pro measured |
| Design Arena: UI components | Not measured | 1246 | 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 | 1243 | Only MiMo-V2.5-Pro measured |
| Design Arena: all code categories | Not measured | 1265 | Only MiMo-V2.5-Pro measured |
| Design Arena: SVG | Not measured | 1181 | Only MiMo-V2.5-Pro measured |
| Design Arena: ASCII art | Not measured | 1156 | Only MiMo-V2.5-Pro measured |
| LiveBench design | 58.4% | Not measured | Only Gemma 4 31B measured |
GeneralLike-for-likePublic boards, shown as published, kept out of the index216
| Benchmark | Gemma 4 31B | MiMo-V2.5-Pro | Leader |
|---|---|---|---|
| LiveBench data analysis | 53.3% | Not measured | Only Gemma 4 31B measured |
| SAGE (Vals AI) | 55.0% | Not measured | Only Gemma 4 31B measured |
| MedScribe (Vals AI) | Not measured | 83.7% | Only MiMo-V2.5-Pro measured |
| IFBench | 75.6% | 79.9% | MiMo-V2.5-Pro |
| Long context reasoning (AA-LCR) | 69.7% | 79.7% | MiMo-V2.5-Pro |
| DTBench | 82.7% | 84.5% | MiMo-V2.5-Pro |
| LMCA | 39.3% | 29.5% | Gemma 4 31B |
| Surface Evolver Bench | 30.6% | Not measured | Only Gemma 4 31B measured |
| LMArena Text | 1443 | 1464 | MiMo-V2.5-Pro |
| LMArena Text: instruction following | 1433 | 1475 | MiMo-V2.5-Pro |
| LMArena Text: creative writing | 1417 | 1438 | MiMo-V2.5-Pro |
| LMArena Vision | 1276 | Not measured | Only Gemma 4 31B measured |
| LMArena Text: English | 1451 | 1477 | MiMo-V2.5-Pro |
| LMArena Text: non-English | 1430 | 1448 | MiMo-V2.5-Pro |
| LMArena Text: Chinese | 1475 | 1507 | MiMo-V2.5-Pro |
| LMArena Text: German | Not measured | 1459 | Only MiMo-V2.5-Pro measured |
| 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 | 1436 | 1487 | MiMo-V2.5-Pro |
| LMArena Text: Spanish | 1442 | 1469 | MiMo-V2.5-Pro |
| LMArena Text: Russian | 1460 | 1449 | Gemma 4 31B |
| LMArena Text: Polish | 1433 | 1464 | MiMo-V2.5-Pro |
| LMArena Text: multi-turn | 1452 | 1476 | MiMo-V2.5-Pro |
| LMArena Text: longer queries | 1446 | 1481 | MiMo-V2.5-Pro |
| LMArena Text: writing and language | 1432 | 1451 | MiMo-V2.5-Pro |
| LMArena Text: entertainment and media | 1405 | 1438 | MiMo-V2.5-Pro |
| LMArena Vision: English | 1271 | Not measured | Only Gemma 4 31B measured |
| LMArena Vision: Chinese | 1326 | Not measured | Only Gemma 4 31B measured |
| LMArena Vision: reading text | 1282 | Not measured | Only Gemma 4 31B measured |
| LMArena Vision: humor | 1280 | Not measured | Only Gemma 4 31B measured |
| LMArena Vision: captioning | 1218 | Not measured | Only Gemma 4 31B measured |
| LMArena Vision: recognizing things | 1226 | Not measured | Only Gemma 4 31B measured |
| LMArena Vision: creative writing | 1285 | Not measured | Only Gemma 4 31B measured |
| LMArena Document | 1425 | Not measured | Only Gemma 4 31B measured |
| Neither model has a result on the other 12 boards here | |||
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | Gemma 4 31B | 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
Gemma 4 31B costs less
Repository review
50K input + 3K output tokens
Gemma 4 31B costs less
Agent loop
200K input + 10K output tokens
Gemma 4 31B 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 | Google Gemma 4 31B | Xiaomi MiMo-V2.5-Pro |
|---|---|---|
| Maker | Xiaomi | |
| Weights | Open weights | Open weights |
| Released | April 2, 2026 | April 22, 2026 |
| Context window | 262K | 1.05M |
| Input $/Mtok | $0.14 | $0.4575 |
| Output $/Mtok | $0.40 | $0.9152 |
| Coding index | 39.0 | 31.6 |
| Independent measurements behind it | 2 | 4 |
| Cost per solved task | Not computable | $0.03 |
| Source | OpenRouter listing | OpenRouter listing |
Questions
Is Gemma 4 31B better than MiMo-V2.5-Pro for coding?
Pick Gemma 4 31B for coding: it scores 39.0 on the Vibe Coding Index against 31.6 for MiMo-V2.5-Pro, though on 2 benchmarks against 4. Pick MiMo-V2.5-Pro for a larger context window (1.05M against 262K tokens) and work measured by LMArena WebDev or AA Coding Index, where it leads.
Which is cheaper, Gemma 4 31B or MiMo-V2.5-Pro?
Gemma 4 31B costs $0.14 per million input tokens and $0.40 per million output tokens; MiMo-V2.5-Pro costs $0.4575 and $0.9152. Cost per solved task is known for MiMo-V2.5-Pro only.
Which has the bigger context window, Gemma 4 31B or MiMo-V2.5-Pro?
MiMo-V2.5-Pro, with 1.05M tokens against 262K.
How many benchmarks is the Gemma 4 31B vs MiMo-V2.5-Pro comparison based on?
41 benchmarks have a result for both models as of October 22, 2026. Gemma 4 31B's coding index rests on 2 independent benchmarks and MiMo-V2.5-Pro's on 4. Where only one model has a result, the page says so instead of naming a winner.