# DeepSeek V4.1 Flash vs Grok 4.3

> DeepSeek V4.1 Flash is the stronger model overall, scoring 50.3 to 43.6 on the Noometry Index.

- Canonical page: https://noometry.com/compare/deepseek-v4-1-flash-vs-grok-4-3
- Last updated: 2026-10-10
- Shared benchmarks: 23

## Summary

- They share 23 benchmarks with published results for both. DeepSeek V4.1 Flash scores higher in 7 categories and Grok 4.3 in 1 category; 8 gaps are clear of the uncertainty.
- The widest gap is in coding, where DeepSeek V4.1 Flash leads 52.8 to 41.3.
- The biggest single-benchmark swing is ProofBench: 54% for DeepSeek V4.1 Flash and 11% for Grok 4.3.
- DeepSeek V4.1 Flash is cheaper at $0.15 / $0.60 per million input/output tokens, against $1.25 / $2.50 for Grok 4.3.
- DeepSeek V4.1 Flash has downloadable open weights; the other is API-only.

## Snapshot

| | DeepSeek V4.1 Flash | Grok 4.3 |
|---|---|---|
| Provider | DeepSeek | xAI |
| Noometry Index | 50.3 | 43.6 |
| Rank | 40 | 80 |
| Context | 1M | 1M |
| Input $/M | $0.15 | $1.25 |
| Output $/M | $0.60 | $2.50 |
| Weights | Open | Proprietary |

## Coding

- DeepSeek V4.1 Flash: 52.8 (#29)
- Grok 4.3: 41.3 (#117)

| Benchmark | DeepSeek V4.1 Flash | Grok 4.3 |
|---|---|---|
| LMArena WebDev | 1620 | 1357 |
| SciCode | 51.9% | 47.3% |
| LMArena Coding | 1511 | 1415 |
| ALE-Bench | 1,092 | 944.17 |
| WeirdML | — | 49.9% |

## Agentic & Tool Use

- DeepSeek V4.1 Flash: —
- Grok 4.3: 27.6 (#93)

| Benchmark | DeepSeek V4.1 Flash | Grok 4.3 |
|---|---|---|
| GDP.pdf | — | 8% |
| LMArena Search | — | 1165 |
| Vending-Bench 2 | — | 35.26 |

## Reasoning

- DeepSeek V4.1 Flash: 46.2 (#39)
- Grok 4.3: 34.9 (#60)

| Benchmark | DeepSeek V4.1 Flash | Grok 4.3 |
|---|---|---|
| NYT Connections (extended) | 89.6% | 55.2% |
| CritPt | 14.3% | 8% |
| LMArena Hard Prompts | 1485 | 1397 |
| DTBench | 89.9% | 90.7% |
| LMCA | 47% | 38.3% |
| Epoch Capabilities Index | 155.01 | 149.13 |
| Chess Puzzles | — | 25% |
| Surface Evolver Bench | 46.3% | — |
| ForecastBench | — | 60.3 |

## Math

- DeepSeek V4.1 Flash: 48.4 (#52)
- Grok 4.3: 45.8 (#60)

| Benchmark | DeepSeek V4.1 Flash | Grok 4.3 |
|---|---|---|
| ProofBench | 54% | 11% |
| LMArena Math | 1493 | 1388 |
| FrontierMath (Tiers 1-3) | — | 42.8% |
| FrontierMath Tier 4 | — | 14.6% |
| OTIS Mock AIME 2024-2025 | — | 93.3% |

## Knowledge

- DeepSeek V4.1 Flash: 41.6 (#97)
- Grok 4.3: 52.5 (#57)

| Benchmark | DeepSeek V4.1 Flash | Grok 4.3 |
|---|---|---|
| LMArena Expert | 1505 | 1385 |
| GPQA Diamond | — | 88.8% |
| SimpleQA Verified | — | 33.2% |

## Multimodal

- DeepSeek V4.1 Flash: —
- Grok 4.3: 31.6 (#96)

| Benchmark | DeepSeek V4.1 Flash | Grok 4.3 |
|---|---|---|
| LMArena Vision | — | 1231 |
| Blueprint-Bench 2 | — | 0% |

## Multilingual

- DeepSeek V4.1 Flash: 55.5 (#29)
- Grok 4.3: 50.7 (#114)

| Benchmark | DeepSeek V4.1 Flash | Grok 4.3 |
|---|---|---|
| LMArena Non-English | 1452 | 1385 |
| LMArena Chinese | 1506 | 1423 |
| LMArena French | 1448 | 1412 |
| LMArena Russian | 1473 | 1399 |
| LMArena Spanish | 1499 | 1402 |
| LMArena German | — | 1395 |
| LMArena Japanese | — | 1379 |
| LMArena Korean | — | 1357 |

## Instruction Following

- DeepSeek V4.1 Flash: 77.6 (#24)
- Grok 4.3: 72.3 (#135)

| Benchmark | DeepSeek V4.1 Flash | Grok 4.3 |
|---|---|---|
| LMArena Instruction Following | 1475 | 1366 |

## Long Context

- DeepSeek V4.1 Flash: 45.2 (#46)
- Grok 4.3: 42.5 (#118)

| Benchmark | DeepSeek V4.1 Flash | Grok 4.3 |
|---|---|---|
| LMArena Longer Query | 1476 | 1393 |

## Writing & Preference

- DeepSeek V4.1 Flash: 65.8 (#43)
- Grok 4.3: 58.6 (#112)

| Benchmark | DeepSeek V4.1 Flash | Grok 4.3 |
|---|---|---|
| LMArena Text | 1465 | 1397 |
| LMArena Creative Writing | 1436 | 1380 |
| LMArena Multi-Turn | 1460 | 1406 |
| EQ-Bench Creative Writing | 1540 | — |
| EQ-Bench 4 | — | 1075 |

## FAQ

### Is DeepSeek V4.1 Flash better than Grok 4.3?

DeepSeek V4.1 Flash is the stronger model overall, scoring 50.3 to 43.6 on the Noometry Index.

### Which is cheaper, DeepSeek V4.1 Flash or Grok 4.3?

DeepSeek V4.1 Flash is cheaper. It lists at $0.15 per million input tokens and $0.60 per million output tokens; Grok 4.3 lists at $1.25 and $2.50.

### Is DeepSeek V4.1 Flash or Grok 4.3 better for coding?

DeepSeek V4.1 Flash scores higher on coding benchmarks: 52.8 versus 41.3 in the Noometry coding category.

### Which has the bigger context window?

Both accept 1M tokens.

### How many benchmarks do DeepSeek V4.1 Flash and Grok 4.3 share?

23 benchmarks have published results for both models. DeepSeek V4.1 Flash has 25 scored results on Noometry and Grok 4.3 has 40.