# DeepSeek V4 Flash vs Ministral 3B

> DeepSeek V4 Flash is the stronger model overall, scoring 53.6 to 25.6 on the Noometry Index. Ministral 3B costs 6.6× less per token, which makes it the better buy when DeepSeek V4 Flash's lead doesn't matter for your workload.

- Canonical page: https://noometry.com/compare/deepseek-v4-flash-vs-ministral-3b
- Last updated: 2026-10-10
- Shared benchmarks: 4

## Summary

- They share 4 benchmarks with published results for both. DeepSeek V4 Flash scores higher in 3 categories and Ministral 3B in 0 categories; 3 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where DeepSeek V4 Flash leads 55.5 to 10.4.
- The biggest single-benchmark swing is GPQA Diamond: 91% for DeepSeek V4 Flash and 25.3% for Ministral 3B.
- Ministral 3B is cheaper at $0.04 / $0.04 per million input/output tokens, against $0.15 / $0.60 for DeepSeek V4 Flash.
- DeepSeek V4 Flash accepts more context: 1M tokens versus 128K.

## Snapshot

| | DeepSeek V4 Flash | Ministral 3B |
|---|---|---|
| Provider | DeepSeek | Mistral AI |
| Noometry Index | 53.6 | 25.6 |
| Rank | 32 | 336 |
| Context | 1M | 128K |
| Input $/M | $0.15 | $0.04 |
| Output $/M | $0.60 | $0.04 |
| Weights | Open | Open |

## Coding

- DeepSeek V4 Flash: 47.6 (#56)
- Ministral 3B: —

| Benchmark | DeepSeek V4 Flash | Ministral 3B |
|---|---|---|
| FrontierCode | 18.8% | — |
| LMArena WebDev | 1581 | — |
| SciCode | 49.9% | — |
| WeirdML | 63% | — |
| LMArena Coding | 1457 | — |
| ALE-Bench | 1,306 | — |

## Reasoning

- DeepSeek V4 Flash: 53.7 (#25)
- Ministral 3B: 17.7 (#278)

| Benchmark | DeepSeek V4 Flash | Ministral 3B |
|---|---|---|
| DTBench | 90.9% | 51.7% |
| LMCA | 41.7% | 5.5% |
| Epoch Capabilities Index | 154.51 | 118.08 |
| ARC-AGI-2 | 61.4% | — |
| SimpleBench | 61.1% | — |
| Kagi LLM Benchmark | 52.2% | — |
| NYT Connections (extended) | 89.6% | — |
| ARC-AGI-1 | 89% | — |
| CritPt | 16.6% | — |
| Chess Puzzles | 33% | — |
| LMArena Hard Prompts | 1443 | — |
| Mystery Game Puzzles | 34% | — |

## Math

- DeepSeek V4 Flash: 60.0 (#31)
- Ministral 3B: 24.5 (#257)

| Benchmark | DeepSeek V4 Flash | Ministral 3B |
|---|---|---|
| FrontierMath (Tiers 1-3) | 57.5% | — |
| FrontierMath Tier 4 | 24.4% | — |
| MathArena Final-Answer Competitions | 76.5% | — |
| OTIS Mock AIME 2024-2025 | 94.4% | — |
| ProofBench | 56% | — |
| LMArena Math | 1427 | — |
| MATH Level 5 | — | 14.4% |

## Knowledge

- DeepSeek V4 Flash: 55.5 (#43)
- Ministral 3B: 10.4 (#297)

| Benchmark | DeepSeek V4 Flash | Ministral 3B |
|---|---|---|
| GPQA Diamond | 91% | 25.3% |
| SimpleQA Verified | 33.6% | — |
| Vectara Hallucination Rate | — | 7.3% |
| LMArena Expert | 1441 | — |

## Multilingual

- DeepSeek V4 Flash: 53.2 (#71)
- Ministral 3B: —

| Benchmark | DeepSeek V4 Flash | Ministral 3B |
|---|---|---|
| LMArena Non-English | 1420 | — |
| LMArena Chinese | 1467 | — |
| LMArena French | 1438 | — |
| LMArena German | 1418 | — |
| LMArena Japanese | 1403 | — |
| LMArena Korean | 1383 | — |
| LMArena Russian | 1428 | — |
| LMArena Spanish | 1437 | — |

## Instruction Following

- DeepSeek V4 Flash: 75.1 (#76)
- Ministral 3B: —

| Benchmark | DeepSeek V4 Flash | Ministral 3B |
|---|---|---|
| LMArena Instruction Following | 1421 | — |

## Long Context

- DeepSeek V4 Flash: 43.8 (#81)
- Ministral 3B: —

| Benchmark | DeepSeek V4 Flash | Ministral 3B |
|---|---|---|
| LMArena Longer Query | 1435 | — |

## Writing & Preference

- DeepSeek V4 Flash: 64.0 (#59)
- Ministral 3B: —

| Benchmark | DeepSeek V4 Flash | Ministral 3B |
|---|---|---|
| LMArena Text | 1432 | — |
| LMArena Creative Writing | 1403 | — |
| EQ-Bench Creative Writing | 1559 | — |
| LMArena Multi-Turn | 1449 | — |

## FAQ

### Is DeepSeek V4 Flash better than Ministral 3B?

DeepSeek V4 Flash is the stronger model overall, scoring 53.6 to 25.6 on the Noometry Index. Ministral 3B costs 6.6× less per token, which makes it the better buy when DeepSeek V4 Flash's lead doesn't matter for your workload.

### Which is cheaper, DeepSeek V4 Flash or Ministral 3B?

Ministral 3B is cheaper. It lists at $0.04 per million input tokens and $0.04 per million output tokens; DeepSeek V4 Flash lists at $0.15 and $0.60.

### Which has the bigger context window?

DeepSeek V4 Flash does, with 1M tokens against 128K.

### How many benchmarks do DeepSeek V4 Flash and Ministral 3B share?

4 benchmarks have published results for both models. DeepSeek V4 Flash has 41 scored results on Noometry and Ministral 3B has 6.