# Deepseek Coder v2 vs Grok 4

> Grok 4 is the stronger model overall, scoring 47.8 to 35.3 on the Noometry Index.

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

## Summary

- They share 17 benchmarks with published results for both. Deepseek Coder v2 scores higher in 0 categories and Grok 4 in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in long context, where Grok 4 leads 63.1 to 37.0.
- Deepseek Coder v2 has downloadable open weights; the other is API-only.

## Snapshot

| | Deepseek Coder v2 | Grok 4 |
|---|---|---|
| Provider | DeepSeek | xAI |
| Noometry Index | 35.3 | 47.8 |
| Rank | 217 | 49 |
| Context | — | — |
| Input $/M | — | — |
| Output $/M | — | — |
| Weights | Open | Proprietary |

## Coding

- Deepseek Coder v2: 37.5 (#178)
- Grok 4: 49.9 (#44)

| Benchmark | Deepseek Coder v2 | Grok 4 |
|---|---|---|
| LMArena Coding | 1251 | 1408 |
| Aider Polyglot | — | 79.6% |
| WeirdML | — | 45.7% |
| BigCodeBench Instruct | 48.2% | — |
| BigCodeBench Complete | 59.7% | — |
| HumanEval+ | 82.3% | — |
| MBPP+ | 75.1% | — |

## Agentic & Tool Use

- Deepseek Coder v2: —
- Grok 4: 32.1 (#62)

| Benchmark | Deepseek Coder v2 | Grok 4 |
|---|---|---|
| Terminal-Bench | — | 27.2% |
| Berkeley Function Calling Leaderboard | — | 63% |
| GDPval | — | 21.1% |
| Cybench | — | 43% |
| DeepResearch Bench | — | 47.3% |
| BALROG | — | 43.6% |
| LMArena Search | — | 1142 |
| METR Time Horizons | — | 66.6% |

## Reasoning

- Deepseek Coder v2: 22.6 (#175)
- Grok 4: 36.9 (#59)

| Benchmark | Deepseek Coder v2 | Grok 4 |
|---|---|---|
| LMArena Hard Prompts | 1207 | 1409 |
| ARC-AGI-2 | — | 16% |
| SimpleBench | — | 60.5% |
| Kagi LLM Benchmark | — | 73.6% |
| ARC-AGI-1 | — | 66.7% |
| Chess Puzzles | — | 28% |
| Epoch Capabilities Index | — | 146.46 |
| ForecastBench | — | 60.9 |
| WinoGrande | 83.7% | — |

## Math

- Deepseek Coder v2: 32.3 (#188)
- Grok 4: 46.5 (#58)

| Benchmark | Deepseek Coder v2 | Grok 4 |
|---|---|---|
| LMArena Math | 1241 | 1422 |
| OTIS Mock AIME 2024-2025 | — | 84% |
| Omni-MATH | — | 60.3% |
| FrontierMath (Feb 2025 set) | — | 19.7% |
| FrontierMath Tier 4 (v1) | — | 2.1% |
| GSM8K | 94.5% | — |

## Knowledge

- Deepseek Coder v2: 31.7 (#209)
- Grok 4: 53.7 (#50)

| Benchmark | Deepseek Coder v2 | Grok 4 |
|---|---|---|
| LMArena Expert | 1181 | 1417 |
| GPQA Diamond | — | 87% |
| MMLU-Pro | — | 85.1% |
| Confabulations | — | 12.4% |
| GPQA (HELM) | — | 72.7% |
| ARC (AI2) Challenge | 64.3% | — |

## Multimodal

- Deepseek Coder v2: —
- Grok 4: 33.1 (#89)

| Benchmark | Deepseek Coder v2 | Grok 4 |
|---|---|---|
| LMArena Vision | — | 1211 |
| GeoBench | — | 45% |

## Multilingual

- Deepseek Coder v2: 36.4 (#235)
- Grok 4: 52.0 (#98)

| Benchmark | Deepseek Coder v2 | Grok 4 |
|---|---|---|
| LMArena Non-English | 1182 | 1403 |
| LMArena Chinese | 1201 | 1428 |
| LMArena French | 1184 | 1418 |
| LMArena German | 1164 | 1429 |
| LMArena Japanese | 1126 | 1396 |
| LMArena Korean | 1104 | 1376 |
| LMArena Russian | 1188 | 1409 |
| LMArena Spanish | 1153 | 1419 |

## Instruction Following

- Deepseek Coder v2: 61.9 (#237)
- Grok 4: 79.4 (#4)

| Benchmark | Deepseek Coder v2 | Grok 4 |
|---|---|---|
| LMArena Instruction Following | 1180 | 1387 |
| IFEval | — | 94.9% |

## Long Context

- Deepseek Coder v2: 37.0 (#219)
- Grok 4: 63.1 (#4)

| Benchmark | Deepseek Coder v2 | Grok 4 |
|---|---|---|
| LMArena Longer Query | 1219 | 1409 |
| Fiction.LiveBench | — | 94.4% |

## Writing & Preference

- Deepseek Coder v2: 38.4 (#249)
- Grok 4: 58.7 (#111)

| Benchmark | Deepseek Coder v2 | Grok 4 |
|---|---|---|
| LMArena Text | 1191 | 1412 |
| LMArena Creative Writing | 1120 | 1397 |
| LMArena Multi-Turn | 1176 | 1416 |
| Short-Story Creative Writing | — | 76.9% |
| WildBench | — | 79.7% |

## FAQ

### Is Deepseek Coder v2 better than Grok 4?

Grok 4 is the stronger model overall, scoring 47.8 to 35.3 on the Noometry Index.

### Is Deepseek Coder v2 or Grok 4 better for coding?

Grok 4 scores higher on coding benchmarks: 49.9 versus 37.5 in the Noometry coding category.

### How many benchmarks do Deepseek Coder v2 and Grok 4 share?

17 benchmarks have published results for both models. Deepseek Coder v2 has 24 scored results on Noometry and Grok 4 has 48.