# DeepSeek V4.1 Flash vs Mistral Medium

> DeepSeek V4.1 Flash is the stronger model overall, scoring 52.8 to 36.3 on the Noometry Index.

- Canonical page: https://noometry.com/compare/deepseek-v4-1-flash-vs-mistral-medium
- Last updated: 2026-10-12
- Shared benchmarks: 27

## Summary

- They share 27 benchmarks with published results for both. DeepSeek V4.1 Flash scores higher in 10 categories and Mistral Medium in 0 categories; 10 gaps are clear of the uncertainty.
- The widest gap is in math, where DeepSeek V4.1 Flash leads 66.7 to 28.1.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 98.3% for DeepSeek V4.1 Flash and 32.2% for Mistral Medium.
- DeepSeek V4.1 Flash is cheaper at $0.15 / $0.60 per million input/output tokens, against $1.50 / $7.50 for Mistral Medium.
- DeepSeek V4.1 Flash accepts more context: 1M tokens versus 262K.

## Snapshot

| | DeepSeek V4.1 Flash | Mistral Medium |
|---|---|---|
| Provider | DeepSeek | Mistral AI |
| Noometry Index | 52.8 | 36.3 |
| Rank | 38 | 218 |
| Context | 1M | 262K |
| Input $/M | $0.15 | $1.50 |
| Output $/M | $0.60 | $7.50 |
| Weights | Open | Open |

## Coding

- DeepSeek V4.1 Flash: 52.9 (#32)
- Mistral Medium: 34.2 (#243)

| Benchmark | DeepSeek V4.1 Flash | Mistral Medium |
|---|---|---|
| SciCode | 51.9% | 40.2% |
| LMArena Coding | 1506 | 1434 |
| ALE-Bench | 1,092 | 763.98 |
| FrontierCode | — | 8% |
| LMArena WebDev | 1619 | — |
| WeirdML | — | 43.7% |

## Agentic & Tool Use

- DeepSeek V4.1 Flash: 31.2 (#69)
- Mistral Medium: 28.3 (#90)

| Benchmark | DeepSeek V4.1 Flash | Mistral Medium |
|---|---|---|
| APEX-Agents | 39.5% | — |
| Berkeley Function Calling Leaderboard | — | 37.7% |
| GDP.pdf | 19.8% | — |

## Reasoning

- DeepSeek V4.1 Flash: 50.2 (#36)
- Mistral Medium: 24.0 (#167)

| Benchmark | DeepSeek V4.1 Flash | Mistral Medium |
|---|---|---|
| CritPt | 14.3% | 0% |
| LMArena Hard Prompts | 1483 | 1426 |
| DTBench | 89.9% | 75.5% |
| LMCA | 47% | 26.1% |
| Surface Evolver Bench | 46.3% | 26.9% |
| Kagi LLM Benchmark | — | 50% |
| NYT Connections (extended) | 89.6% | — |
| Mystery Game Puzzles | 43% | — |
| Epoch Capabilities Index | 154.9 | — |

## Math

- DeepSeek V4.1 Flash: 66.7 (#25)
- Mistral Medium: 28.1 (#245)

| Benchmark | DeepSeek V4.1 Flash | Mistral Medium |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 98.3% | 32.2% |
| ProofBench | 54% | 9% |
| LMArena Math | 1477 | 1408 |
| FrontierMath (Tiers 1-3) | 67.4% | — |
| FrontierMath Tier 4 | 26.8% | — |
| MATH Level 5 | — | 81.6% |
| FrontierMath (Feb 2025 set) | — | 0.3% |

## Knowledge

- DeepSeek V4.1 Flash: 57.9 (#38)
- Mistral Medium: 25.0 (#265)

| Benchmark | DeepSeek V4.1 Flash | Mistral Medium |
|---|---|---|
| GPQA Diamond | 89.8% | 59.5% |
| LMArena Expert | 1506 | 1408 |
| Humanity's Last Exam | — | 4.5% |
| Vectara Hallucination Rate | — | 22.7% |

## Multimodal

- DeepSeek V4.1 Flash: 39.1 (#61)
- Mistral Medium: 35.3 (#88)

| Benchmark | DeepSeek V4.1 Flash | Mistral Medium |
|---|---|---|
| LMArena Vision | 1277 | 1172 |
| Furniture Assembly | 34.2% | — |

## Multilingual

- DeepSeek V4.1 Flash: 55.0 (#35)
- Mistral Medium: 52.1 (#91)

| Benchmark | DeepSeek V4.1 Flash | Mistral Medium |
|---|---|---|
| LMArena Non-English | 1448 | 1408 |
| LMArena Chinese | 1497 | 1447 |
| LMArena French | 1452 | 1459 |
| LMArena German | 1484 | 1432 |
| LMArena Japanese | 1412 | 1378 |
| LMArena Korean | 1452 | 1380 |
| LMArena Russian | 1471 | 1411 |
| LMArena Spanish | 1459 | 1433 |

## Instruction Following

- DeepSeek V4.1 Flash: 77.3 (#26)
- Mistral Medium: 73.7 (#116)

| Benchmark | DeepSeek V4.1 Flash | Mistral Medium |
|---|---|---|
| LMArena Instruction Following | 1474 | 1398 |

## Long Context

- DeepSeek V4.1 Flash: 45.2 (#47)
- Mistral Medium: 42.9 (#114)

| Benchmark | DeepSeek V4.1 Flash | Mistral Medium |
|---|---|---|
| LMArena Longer Query | 1475 | 1406 |

## Writing & Preference

- DeepSeek V4.1 Flash: 65.4 (#48)
- Mistral Medium: 60.0 (#103)

| Benchmark | DeepSeek V4.1 Flash | Mistral Medium |
|---|---|---|
| LMArena Text | 1462 | 1424 |
| LMArena Creative Writing | 1435 | 1391 |
| LMArena Multi-Turn | 1457 | 1418 |
| Short-Story Creative Writing | — | 77.3% |
| EQ-Bench Creative Writing | 1540 | — |

## FAQ

### Is DeepSeek V4.1 Flash better than Mistral Medium?

DeepSeek V4.1 Flash is the stronger model overall, scoring 52.8 to 36.3 on the Noometry Index.

### Which is cheaper, DeepSeek V4.1 Flash or Mistral Medium?

DeepSeek V4.1 Flash is cheaper. It lists at $0.15 per million input tokens and $0.60 per million output tokens; Mistral Medium lists at $1.50 and $7.50.

### Is DeepSeek V4.1 Flash or Mistral Medium better for coding?

DeepSeek V4.1 Flash scores higher on coding benchmarks: 52.9 versus 34.2 in the Noometry coding category.

### Which has the bigger context window?

DeepSeek V4.1 Flash does, with 1M tokens against 262K.

### How many benchmarks do DeepSeek V4.1 Flash and Mistral Medium share?

27 benchmarks have published results for both models. DeepSeek V4.1 Flash has 37 scored results on Noometry and Mistral Medium has 36.