MiMo-V2.5-Pro vs MiMo-V2.6-Flash
MiMo-V2.6-Flash leads MiMo-V2.5-Pro on coding, 54.7 to 31.6 on the Vibe Coding Index, and MiMo-V2.6-Flash costs 3.3x less per output token ($0.28 vs $0.9152 per million). Updated October 22, 2026; 47 benchmarks have a result for both.
MiMo-V2.5-Pro or MiMo-V2.6-Flash: which to pick
Pick MiMo-V2.6-Flash for coding: it scores 54.7 on the Vibe Coding Index against 31.6 for MiMo-V2.5-Pro, though on 3 benchmarks against 4.
Pick MiMo-V2.5-Pro for work measured by LMArena Text: coding or Humanity's Last Exam (Artificial Analysis run), where it leads.
- Shared results
- 47
- MiMo-V2.5-Pro only
- 24
- MiMo-V2.6-Flash only
- 19
- 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
MiMo-V2.6-Flash
MiMo-V2.6-Flash scores 54.7 against 31.6.
4 and 3 benchmarks behind the scores
Intelligence
AA Intelligence Index
MiMo-V2.6-Flash
MiMo-V2.6-Flash leads, 37.9 against 26.0.
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, 1266.
Directional only
General
Vals Index
No call
Only MiMo-V2.6-Flash has a Vals Index result, 54.8%.
Directional only
Long documents
Listed context window
No call
Both take 1.05M tokens.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
MiMo-V2.6-Flash
MiMo-V2.5-Pro $0.0009, MiMo-V2.6-Flash $0.0003 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
- MiMo-V2.5-Pro 31.6 from 4 benchmarks, MiMo-V2.6-Flash 54.7 from 3 benchmarks.
- Widest category
- Coding, 17.5 points apart on average over 11 shared benchmarks.
- Widest single gap
- Vibe Code Bench (Vals AI): MiMo-V2.5-Pro 34.1% against MiMo-V2.6-Flash 79.0%, 44.8 points apart on a 0 to 100 scale.
- Output price
- MiMo-V2.5-Pro $0.9152 against MiMo-V2.6-Flash $0.28 per million output tokens, 3.3x apart.
- Context window
- Both take 1.05M tokens.
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 | MiMo-V2.5-Pro | Bars | MiMo-V2.6-Flash | Ties |
|---|---|---|---|---|
| Coding | 1 | 10 | ||
| Intelligence | 0 | 1 | ||
| Agentic work | 0 | 4 | ||
| Reasoning | 4 | 0 | ||
| Math | 2 | 1 | ||
| Knowledge | 4 | 1 | ||
| General | 17 | 2 | ||
| All shared boards | 28 | 19 |
CodingLike-for-likeSix of these boards feed the Vibe Coding Index110
| Benchmark | MiMo-V2.5-Pro | MiMo-V2.6-Flash | Leader |
|---|---|---|---|
| LMArena WebDev | 1480 | 1634 | MiMo-V2.6-Flash |
| AA Coding Index | 60.2 | Not measured | Only MiMo-V2.5-Pro measured |
| LiveCodeBench (Vals AI) | 81.3% | Not measured | Only MiMo-V2.5-Pro measured |
| SWE-bench Verified (Vals AI) | 74.0% | Not measured | Only MiMo-V2.5-Pro measured |
| Terminal-Bench 4 (Vals AI) | 0.5% | 24.2% | MiMo-V2.6-Flash |
| Vibe Code Bench (Vals AI) | 34.1% | 79.0% | MiMo-V2.6-Flash |
| Code migration (Vals AI) | 21.6% | 40.9% | MiMo-V2.6-Flash |
| ProgramBench (Vals AI) | Not measured | 0.5% | Only MiMo-V2.6-Flash measured |
| IOI (Vals AI) | Not measured | 47.7% | Only MiMo-V2.6-Flash measured |
| Terminal-Bench 2.1 (Vals AI) | 57.3% | 76.4% | MiMo-V2.6-Flash |
| CyberBench (Vals AI) | Not measured | 75.4% | Only MiMo-V2.6-Flash measured |
| SciCode | 50.6% | 51.3% | MiMo-V2.6-Flash |
| Terminal-Bench Hard (Artificial Analysis run) | 43.2% | Not measured | Only MiMo-V2.5-Pro measured |
| ALE-Bench | 899.8 | 1244.2 | MiMo-V2.6-Flash |
| LMArena Text: coding | 1501 | 1499 | MiMo-V2.5-Pro |
| LMArena Text: software and IT | 1495 | 1497 | MiMo-V2.6-Flash |
| LMArena WebDev: React | 1469 | 1648 | MiMo-V2.6-Flash |
| LMArena WebDev: HTML | 1473 | 1641 | MiMo-V2.6-Flash |
| LMArena WebDev: from an image | Not measured | 1596 | Only MiMo-V2.6-Flash measured |
| Neither model has a result on the other 24 boards here | |||
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index01
| Benchmark | MiMo-V2.5-Pro | MiMo-V2.6-Flash | Leader |
|---|---|---|---|
| AA Intelligence Index | 26.0 | 37.9 | MiMo-V2.6-Flash |
| Neither model has a result on the other 1 board here | |||
Agentic workLike-for-likePublic boards, shown as published, kept out of the index04
| Benchmark | MiMo-V2.5-Pro | MiMo-V2.6-Flash | Leader |
|---|---|---|---|
| AA Agentic Index | 21.3 | Not measured | Only MiMo-V2.5-Pro measured |
| SREBench (Vals AI) | Not measured | 4.2% | Only MiMo-V2.6-Flash measured |
| Finance Agent (Vals AI) | 41.5% | 56.3% | MiMo-V2.6-Flash |
| Tax Agent Bench (Vals AI) | 11.5% | 28.5% | MiMo-V2.6-Flash |
| Legal Research Bench (Vals AI) | 15.9% | 38.0% | MiMo-V2.6-Flash |
| Harvey's Legal Agent Benchmark (Vals AI) | Not measured | 13.3% | Only MiMo-V2.6-Flash measured |
| Excel Modeling Benchmark (Vals AI) | 55.2% | 65.5% | MiMo-V2.6-Flash |
| BioMysteryBench (Vals AI) | Not measured | 69.3% | Only MiMo-V2.6-Flash measured |
| Terminal-Bench-Science (Vals AI) | Not measured | 4.3% | Only MiMo-V2.6-Flash measured |
| Vals Science Index | Not measured | 31.7% | Only MiMo-V2.6-Flash measured |
| τ²-Bench Telecom | 94.2% | Not measured | Only MiMo-V2.5-Pro measured |
| Neither model has a result on the other 31 boards here | |||
ReasoningLike-for-likePublic boards, shown as published, kept out of the index40
| Benchmark | MiMo-V2.5-Pro | MiMo-V2.6-Flash | Leader |
|---|---|---|---|
| GPQA Diamond | 82.6% | Not measured | Only MiMo-V2.5-Pro measured |
| MysteryMechanism (Vals AI) | Not measured | 21.6% | Only MiMo-V2.6-Flash measured |
| GPQA Diamond (Artificial Analysis run) | 86.6% | Not measured | Only MiMo-V2.5-Pro measured |
| Humanity's Last Exam (Artificial Analysis run) | 35.7% | 35.1% | MiMo-V2.5-Pro |
| LMArena Text: hard prompts | 1486 | 1475 | MiMo-V2.5-Pro |
| LMArena Text: expert | 1501 | 1496 | MiMo-V2.5-Pro |
| LMArena Text: hard prompts in English | 1494 | 1493 | MiMo-V2.5-Pro |
| LMArena Vision: diagrams | Not measured | 1276 | Only MiMo-V2.6-Flash measured |
| LMArena Vision: homework | Not measured | 1283 | Only MiMo-V2.6-Flash measured |
| Neither model has a result on the other 14 boards here | |||
MathLike-for-likePublic boards, shown as published, kept out of the index21
| Benchmark | MiMo-V2.5-Pro | MiMo-V2.6-Flash | Leader |
|---|---|---|---|
| ProofBench | 22.0% | 63.0% | MiMo-V2.6-Flash |
| LMArena Text: math | 1481 | 1476 | MiMo-V2.5-Pro |
| LMArena Text: mathematical work | 1491 | 1478 | MiMo-V2.5-Pro |
| Neither model has a result on the other 11 boards here | |||
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index41
| Benchmark | MiMo-V2.5-Pro | MiMo-V2.6-Flash | Leader |
|---|---|---|---|
| MMLU Pro | 84.6% | Not measured | Only MiMo-V2.5-Pro measured |
| LegalBench (Vals AI) | 77.1% | Not measured | Only MiMo-V2.5-Pro measured |
| CorpFin (Vals AI) | 61.4% | Not measured | Only MiMo-V2.5-Pro measured |
| TaxEval (Vals AI) | 73.8% | Not measured | Only MiMo-V2.5-Pro measured |
| MedCode (Vals AI) | 32.5% | 41.1% | MiMo-V2.6-Flash |
| LMArena Text: science | 1482 | 1462 | MiMo-V2.5-Pro |
| LMArena Text: medicine and healthcare | 1473 | 1439 | MiMo-V2.5-Pro |
| LMArena Text: legal and government | 1474 | 1441 | MiMo-V2.5-Pro |
| LMArena Text: business and finance | 1466 | 1454 | MiMo-V2.5-Pro |
| Neither model has a result on the other 11 boards here | |||
Web designDirectional onlyPublic boards, shown as published, kept out of the index8 boards, none shared
| Benchmark | MiMo-V2.5-Pro | MiMo-V2.6-Flash | Leader |
|---|---|---|---|
| Design Arena: websites | 1266 | Not measured | Only MiMo-V2.5-Pro measured |
| Design Arena: UI components | 1246 | Not measured | Only MiMo-V2.5-Pro measured |
| Design Arena: game dev | 1261 | Not measured | Only MiMo-V2.5-Pro measured |
| Design Arena: data viz | 1257 | Not measured | Only MiMo-V2.5-Pro measured |
| Design Arena: 3D | 1243 | Not measured | Only MiMo-V2.5-Pro measured |
| Design Arena: all code categories | 1265 | Not measured | Only MiMo-V2.5-Pro measured |
| Design Arena: SVG | 1182 | Not measured | Only MiMo-V2.5-Pro measured |
| Design Arena: ASCII art | 1156 | Not measured | 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 index172
| Benchmark | MiMo-V2.5-Pro | MiMo-V2.6-Flash | Leader |
|---|---|---|---|
| Vals Index | Not measured | 54.8% | Only MiMo-V2.6-Flash measured |
| Public Benefits Bench (Vals AI) | Not measured | 67.6% | Only MiMo-V2.6-Flash measured |
| SAGE (Vals AI) | Not measured | 43.5% | Only MiMo-V2.6-Flash measured |
| MedScribe (Vals AI) | 83.7% | 85.3% | MiMo-V2.6-Flash |
| IFBench | 79.9% | Not measured | Only MiMo-V2.5-Pro measured |
| Long context reasoning (AA-LCR) | 79.7% | 74.3% | MiMo-V2.5-Pro |
| DTBench | 84.5% | Not measured | Only MiMo-V2.5-Pro measured |
| LMCA | 29.5% | Not measured | Only MiMo-V2.5-Pro measured |
| LMArena Text | 1464 | 1453 | MiMo-V2.5-Pro |
| LMArena Text: instruction following | 1475 | 1452 | MiMo-V2.5-Pro |
| LMArena Text: creative writing | 1438 | 1403 | MiMo-V2.5-Pro |
| LMArena Vision | Not measured | 1256 | Only MiMo-V2.6-Flash measured |
| LMArena Text: English | 1477 | 1476 | MiMo-V2.5-Pro |
| LMArena Text: non-English | 1448 | 1431 | MiMo-V2.5-Pro |
| LMArena Text: Chinese | 1507 | 1509 | MiMo-V2.6-Flash |
| LMArena Text: German | 1459 | 1458 | MiMo-V2.5-Pro |
| LMArena Text: Japanese | 1410 | 1359 | MiMo-V2.5-Pro |
| LMArena Text: Korean | 1435 | 1425 | MiMo-V2.5-Pro |
| LMArena Text: French | 1487 | 1480 | MiMo-V2.5-Pro |
| LMArena Text: Spanish | 1469 | 1453 | MiMo-V2.5-Pro |
| LMArena Text: Russian | 1449 | 1409 | MiMo-V2.5-Pro |
| LMArena Text: Polish | 1464 | 1455 | MiMo-V2.5-Pro |
| LMArena Text: multi-turn | 1476 | 1447 | MiMo-V2.5-Pro |
| LMArena Text: longer queries | 1481 | 1459 | MiMo-V2.5-Pro |
| LMArena Text: writing and language | 1451 | 1424 | MiMo-V2.5-Pro |
| LMArena Text: entertainment and media | 1438 | 1415 | MiMo-V2.5-Pro |
| LMArena Vision: English | Not measured | 1269 | Only MiMo-V2.6-Flash measured |
| LMArena Vision: reading text | Not measured | 1270 | Only MiMo-V2.6-Flash measured |
| LMArena Vision: creative writing | Not measured | 1231 | Only MiMo-V2.6-Flash measured |
| EQ-Bench 4 | 1208 | Not measured | Only MiMo-V2.5-Pro measured |
| Neither model has a result on the other 18 boards here | |||
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | MiMo-V2.5-Pro | MiMo-V2.6-Flash | 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
MiMo-V2.6-Flash costs less
Repository review
50K input + 3K output tokens
MiMo-V2.6-Flash costs less
Agent loop
200K input + 10K output tokens
MiMo-V2.6-Flash 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 | Xiaomi MiMo-V2.5-Pro | Xiaomi MiMo-V2.6-Flash |
|---|---|---|
| Maker | Xiaomi | Xiaomi |
| Weights | Open weights | Open weights |
| Released | April 22, 2026 | September 21, 2026 |
| Context window | 1.05M | 1.05M |
| Input $/Mtok | $0.4575 | $0.14 |
| Output $/Mtok | $0.9152 | $0.28 |
| Coding index | 31.6 | 54.7 |
| Independent measurements behind it | 4 | 3 |
| Cost per solved task | $0.03 | Not computable |
| Source | OpenRouter listing | OpenRouter listing |
Questions
Is MiMo-V2.5-Pro better than MiMo-V2.6-Flash for coding?
Pick MiMo-V2.6-Flash for coding: it scores 54.7 on the Vibe Coding Index against 31.6 for MiMo-V2.5-Pro, though on 3 benchmarks against 4. Pick MiMo-V2.5-Pro for work measured by LMArena Text: coding or Humanity's Last Exam (Artificial Analysis run), where it leads.
Which is cheaper, MiMo-V2.5-Pro or MiMo-V2.6-Flash?
MiMo-V2.5-Pro costs $0.4575 per million input tokens and $0.9152 per million output tokens; MiMo-V2.6-Flash costs $0.14 and $0.28. Cost per solved task is known for MiMo-V2.5-Pro only.
Which has the bigger context window, MiMo-V2.5-Pro or MiMo-V2.6-Flash?
Neither: both take 1.05M tokens.
How many benchmarks is the MiMo-V2.5-Pro vs MiMo-V2.6-Flash comparison based on?
47 benchmarks have a result for both models as of October 22, 2026. MiMo-V2.5-Pro's coding index rests on 4 independent benchmarks and MiMo-V2.6-Flash's on 3. Where only one model has a result, the page says so instead of naming a winner.