# Deepseek Coder v2 vs Grok 4.1

> Grok 4.1 is the stronger model overall, scoring 40.8 to 35.3 on the Noometry Index.

- Canonical page: https://noometry.com/compare/deepseek-coder-v2-vs-grok-4-1
- Last updated: 2026-10-10
- Shared benchmarks: 17

## Summary

- They share 17 benchmarks with published results for both. Deepseek Coder v2 scores higher in 1 category and Grok 4.1 in 7 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where Grok 4.1 leads 62.6 to 38.4.
- Deepseek Coder v2 has downloadable open weights; the other is API-only.

## Snapshot

| | Deepseek Coder v2 | Grok 4.1 |
|---|---|---|
| Provider | DeepSeek | xAI |
| Noometry Index | 35.3 | 40.8 |
| Rank | 217 | 135 |
| Context | — | — |
| Input $/M | — | — |
| Output $/M | — | — |
| Weights | Open | Proprietary |

## Coding

- Deepseek Coder v2: 37.5 (#178)
- Grok 4.1: 33.4 (#246)

| Benchmark | Deepseek Coder v2 | Grok 4.1 |
|---|---|---|
| LMArena Coding | 1251 | 1445 |
| LMArena WebDev | — | 1214 |
| BigCodeBench Instruct | 48.2% | — |
| BigCodeBench Complete | 59.7% | — |
| HumanEval+ | 82.3% | — |
| MBPP+ | 75.1% | — |

## Agentic & Tool Use

- Deepseek Coder v2: —
- Grok 4.1: 33.8 (#47)

| Benchmark | Deepseek Coder v2 | Grok 4.1 |
|---|---|---|
| Cybench | — | 39% |

## Reasoning

- Deepseek Coder v2: 22.6 (#175)
- Grok 4.1: 28.3 (#91)

| Benchmark | Deepseek Coder v2 | Grok 4.1 |
|---|---|---|
| LMArena Hard Prompts | 1207 | 1436 |
| WinoGrande | 83.7% | — |

## Math

- Deepseek Coder v2: 32.3 (#188)
- Grok 4.1: 36.2 (#124)

| Benchmark | Deepseek Coder v2 | Grok 4.1 |
|---|---|---|
| LMArena Math | 1241 | 1422 |
| GSM8K | 94.5% | — |

## Knowledge

- Deepseek Coder v2: 31.7 (#209)
- Grok 4.1: 38.9 (#129)

| Benchmark | Deepseek Coder v2 | Grok 4.1 |
|---|---|---|
| LMArena Expert | 1181 | 1417 |
| ARC (AI2) Challenge | 64.3% | — |

## Multilingual

- Deepseek Coder v2: 36.4 (#235)
- Grok 4.1: 53.6 (#64)

| Benchmark | Deepseek Coder v2 | Grok 4.1 |
|---|---|---|
| LMArena Non-English | 1182 | 1425 |
| LMArena Chinese | 1201 | 1464 |
| LMArena French | 1184 | 1448 |
| LMArena German | 1164 | 1447 |
| LMArena Japanese | 1126 | 1396 |
| LMArena Korean | 1104 | 1407 |
| LMArena Russian | 1188 | 1434 |
| LMArena Spanish | 1153 | 1438 |

## Instruction Following

- Deepseek Coder v2: 61.9 (#237)
- Grok 4.1: 74.0 (#106)

| Benchmark | Deepseek Coder v2 | Grok 4.1 |
|---|---|---|
| LMArena Instruction Following | 1180 | 1400 |

## Long Context

- Deepseek Coder v2: 37.0 (#219)
- Grok 4.1: 43.2 (#96)

| Benchmark | Deepseek Coder v2 | Grok 4.1 |
|---|---|---|
| LMArena Longer Query | 1219 | 1416 |

## Writing & Preference

- Deepseek Coder v2: 38.4 (#249)
- Grok 4.1: 62.6 (#71)

| Benchmark | Deepseek Coder v2 | Grok 4.1 |
|---|---|---|
| LMArena Text | 1191 | 1437 |
| LMArena Creative Writing | 1120 | 1411 |
| LMArena Multi-Turn | 1176 | 1438 |

## FAQ

### Is Deepseek Coder v2 better than Grok 4.1?

Grok 4.1 is the stronger model overall, scoring 40.8 to 35.3 on the Noometry Index.

### Is Deepseek Coder v2 or Grok 4.1 better for coding?

Deepseek Coder v2 scores higher on coding benchmarks: 37.5 versus 33.4 in the Noometry coding category.

### How many benchmarks do Deepseek Coder v2 and Grok 4.1 share?

17 benchmarks have published results for both models. Deepseek Coder v2 has 24 scored results on Noometry and Grok 4.1 has 19.