# DeepSeek V4 Flash vs MiMo-V2.5-Pro

> DeepSeek V4 Flash is the stronger model overall, scoring 53.6 to 44.8 on the Noometry Index.

- Canonical page: https://noometry.com/compare/deepseek-v4-flash-vs-mimo-v2-5-pro
- Last updated: 2026-10-10
- Shared benchmarks: 26

## Summary

- They share 26 benchmarks with published results for both. DeepSeek V4 Flash scores higher in 4 categories and MiMo-V2.5-Pro in 4 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where DeepSeek V4 Flash leads 53.7 to 25.9.
- The biggest single-benchmark swing is NYT Connections (extended): 89.6% for DeepSeek V4 Flash and 34.4% for MiMo-V2.5-Pro.
- DeepSeek V4 Flash is cheaper at $0.15 / $0.60 per million input/output tokens, against $0.43 / $0.87 for MiMo-V2.5-Pro.
- MiMo-V2.5-Pro accepts more context: 1.05M tokens versus 1M.

## Snapshot

| | DeepSeek V4 Flash | MiMo-V2.5-Pro |
|---|---|---|
| Provider | DeepSeek | Xiaomi |
| Noometry Index | 53.6 | 44.8 |
| Rank | 32 | 70 |
| Context | 1M | 1.05M |
| Input $/M | $0.15 | $0.43 |
| Output $/M | $0.60 | $0.87 |
| Weights | Open | Open |

## Coding

- DeepSeek V4 Flash: 47.6 (#56)
- MiMo-V2.5-Pro: 47.0 (#58)

| Benchmark | DeepSeek V4 Flash | MiMo-V2.5-Pro |
|---|---|---|
| LMArena WebDev | 1581 | 1477 |
| SciCode | 49.9% | 50.2% |
| LMArena Coding | 1457 | 1502 |
| ALE-Bench | 1,306 | 899.8 |
| FrontierCode | 18.8% | — |
| WeirdML | 63% | — |

## Reasoning

- DeepSeek V4 Flash: 53.7 (#25)
- MiMo-V2.5-Pro: 25.9 (#124)

| Benchmark | DeepSeek V4 Flash | MiMo-V2.5-Pro |
|---|---|---|
| NYT Connections (extended) | 89.6% | 34.4% |
| CritPt | 16.6% | 4% |
| LMArena Hard Prompts | 1443 | 1487 |
| DTBench | 90.9% | 84.5% |
| LMCA | 41.7% | 29.5% |
| ARC-AGI-2 | 61.4% | — |
| SimpleBench | 61.1% | — |
| Kagi LLM Benchmark | 52.2% | — |
| ARC-AGI-1 | 89% | — |
| Chess Puzzles | 33% | — |
| Mystery Game Puzzles | 34% | — |
| Epoch Capabilities Index | 154.51 | — |

## Math

- DeepSeek V4 Flash: 60.0 (#31)
- MiMo-V2.5-Pro: 38.6 (#84)

| Benchmark | DeepSeek V4 Flash | MiMo-V2.5-Pro |
|---|---|---|
| ProofBench | 56% | 22% |
| LMArena Math | 1427 | 1480 |
| FrontierMath (Tiers 1-3) | 57.5% | — |
| FrontierMath Tier 4 | 24.4% | — |
| MathArena Final-Answer Competitions | 76.5% | — |
| OTIS Mock AIME 2024-2025 | 94.4% | — |

## Knowledge

- DeepSeek V4 Flash: 55.5 (#43)
- MiMo-V2.5-Pro: 41.5 (#100)

| Benchmark | DeepSeek V4 Flash | MiMo-V2.5-Pro |
|---|---|---|
| LMArena Expert | 1441 | 1501 |
| GPQA Diamond | 91% | — |
| SimpleQA Verified | 33.6% | — |

## Multilingual

- DeepSeek V4 Flash: 53.2 (#71)
- MiMo-V2.5-Pro: 55.3 (#33)

| Benchmark | DeepSeek V4 Flash | MiMo-V2.5-Pro |
|---|---|---|
| LMArena Non-English | 1420 | 1449 |
| LMArena Chinese | 1467 | 1507 |
| LMArena French | 1438 | 1489 |
| LMArena German | 1418 | 1455 |
| LMArena Japanese | 1403 | 1418 |
| LMArena Korean | 1383 | 1436 |
| LMArena Russian | 1428 | 1450 |
| LMArena Spanish | 1437 | 1470 |

## Instruction Following

- DeepSeek V4 Flash: 75.1 (#76)
- MiMo-V2.5-Pro: 77.6 (#21)

| Benchmark | DeepSeek V4 Flash | MiMo-V2.5-Pro |
|---|---|---|
| LMArena Instruction Following | 1421 | 1477 |

## Long Context

- DeepSeek V4 Flash: 43.8 (#81)
- MiMo-V2.5-Pro: 45.4 (#38)

| Benchmark | DeepSeek V4 Flash | MiMo-V2.5-Pro |
|---|---|---|
| LMArena Longer Query | 1435 | 1482 |

## Writing & Preference

- DeepSeek V4 Flash: 64.0 (#59)
- MiMo-V2.5-Pro: 65.5 (#47)

| Benchmark | DeepSeek V4 Flash | MiMo-V2.5-Pro |
|---|---|---|
| LMArena Text | 1432 | 1465 |
| LMArena Creative Writing | 1403 | 1440 |
| EQ-Bench Creative Writing | 1559 | 1493 |
| LMArena Multi-Turn | 1449 | 1477 |
| EQ-Bench 4 | — | 1208 |

## FAQ

### Is DeepSeek V4 Flash better than MiMo-V2.5-Pro?

DeepSeek V4 Flash is the stronger model overall, scoring 53.6 to 44.8 on the Noometry Index.

### Which is cheaper, DeepSeek V4 Flash or MiMo-V2.5-Pro?

DeepSeek V4 Flash is cheaper. It lists at $0.15 per million input tokens and $0.60 per million output tokens; MiMo-V2.5-Pro lists at $0.43 and $0.87.

### Is DeepSeek V4 Flash or MiMo-V2.5-Pro better for coding?

They score almost the same on coding (47.6 vs 47.0); test both on your own repository before choosing.

### Which has the bigger context window?

MiMo-V2.5-Pro does, with 1.05M tokens against 1M.

### How many benchmarks do DeepSeek V4 Flash and MiMo-V2.5-Pro share?

26 benchmarks have published results for both models. DeepSeek V4 Flash has 41 scored results on Noometry and MiMo-V2.5-Pro has 27.