# DeepSeek-V3 vs Grok-3 mini

> Grok-3 mini is the stronger model overall, scoring 41.0 to 39.2 on the Noometry Index.

- Canonical page: https://noometry.com/compare/deepseek-v3-vs-grok-3-mini
- Last updated: 2026-10-10
- Shared benchmarks: 33

## Summary

- They share 33 benchmarks with published results for both. DeepSeek-V3 scores higher in 4 categories and Grok-3 mini in 4 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in math, where Grok-3 mini leads 40.3 to 30.4.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 37.8% for DeepSeek-V3 and 77.8% for Grok-3 mini.
- DeepSeek-V3 has downloadable open weights; the other is API-only.

## Snapshot

| | DeepSeek-V3 | Grok-3 mini |
|---|---|---|
| Provider | DeepSeek | xAI |
| Noometry Index | 39.2 | 41.0 |
| Rank | 157 | 130 |
| Context | 164K | — |
| Input $/M | $0.24 | — |
| Output $/M | $0.90 | — |
| Weights | Open | Proprietary |

## Coding

- DeepSeek-V3: 41.9 (#99)
- Grok-3 mini: 40.4 (#126)

| Benchmark | DeepSeek-V3 | Grok-3 mini |
|---|---|---|
| Aider Polyglot | 55.1% | 49.3% |
| WeirdML | 36.1% | 42.6% |
| LMArena Coding | 1368 | 1380 |
| SciCode | 35.8% | — |
| BigCodeBench Instruct | 50% | — |
| LiveBench Coding | 70.9% | — |
| BigCodeBench Complete | 62.2% | — |
| HumanEval+ | 86.6% | — |
| MBPP+ | 73% | — |

## Agentic & Tool Use

- DeepSeek-V3: —
- Grok-3 mini: —

| Benchmark | DeepSeek-V3 | Grok-3 mini |
|---|---|---|
| METR Time Horizons | 49.6% | — |

## Reasoning

- DeepSeek-V3: 20.0 (#226)
- Grok-3 mini: 13.8 (#323)

| Benchmark | DeepSeek-V3 | Grok-3 mini |
|---|---|---|
| Kagi LLM Benchmark | 52.3% | 61.3% |
| LMArena Hard Prompts | 1365 | 1374 |
| Epoch Capabilities Index | 135.95 | 140.35 |
| ARC-AGI-2 | — | 0.4% |
| SimpleBench | 27.2% | — |
| ARC-AGI-1 | — | 16.5% |
| CritPt | 0% | — |
| LiveBench Reasoning | 65.8% | — |
| DTBench | 64.8% | — |
| LiveBench Data Analysis | 60.9% | — |
| LMCA | 15.5% | — |
| BIG-Bench Hard | 87.5% | — |
| ForecastBench | 59.1 | — |
| HellaSwag | 88.9% | — |
| LiveBench | 66.9% | — |
| PIQA | 84.7% | — |
| WinoGrande | 85.2% | — |

## Math

- DeepSeek-V3: 30.4 (#208)
- Grok-3 mini: 40.3 (#78)

| Benchmark | DeepSeek-V3 | Grok-3 mini |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 37.8% | 77.8% |
| Omni-MATH | 40.3% | 31.8% |
| LMArena Math | 1373 | 1386 |
| MATH Level 5 | 75.5% | 90.9% |
| FrontierMath (Feb 2025 set) | 1.7% | 5.9% |
| LiveBench Math | 73.5% | — |

## Knowledge

- DeepSeek-V3: 37.5 (#147)
- Grok-3 mini: 46.3 (#75)

| Benchmark | DeepSeek-V3 | Grok-3 mini |
|---|---|---|
| GPQA Diamond | 67.6% | 76.3% |
| MMLU-Pro | 72.3% | 79.9% |
| Confabulations | 26.1% | 10.8% |
| GPQA (HELM) | 53.8% | 67.5% |
| LMArena Expert | 1352 | 1395 |
| Vectara Hallucination Rate | 6.1% | — |
| ARC (AI2) Challenge | 95.3% | — |
| MMLU | 87.2% | — |
| TriviaQA | 82.9% | — |

## Multilingual

- DeepSeek-V3: 48.8 (#138)
- Grok-3 mini: 48.3 (#141)

| Benchmark | DeepSeek-V3 | Grok-3 mini |
|---|---|---|
| LMArena Non-English | 1358 | 1352 |
| LMArena Chinese | 1390 | 1387 |
| LMArena French | 1386 | 1357 |
| LMArena German | 1374 | 1350 |
| LMArena Japanese | 1333 | 1343 |
| LMArena Korean | 1321 | 1336 |
| LMArena Russian | 1372 | 1352 |
| LMArena Spanish | 1358 | 1379 |

## Instruction Following

- DeepSeek-V3: 73.0 (#125)
- Grok-3 mini: 78.7 (#8)

| Benchmark | DeepSeek-V3 | Grok-3 mini |
|---|---|---|
| IFEval | 83.2% | 95.1% |
| LMArena Instruction Following | 1345 | 1356 |
| LiveBench Instruction Following | 81.5% | — |

## Long Context

- DeepSeek-V3: 34.0 (#248)
- Grok-3 mini: 41.0 (#142)

| Benchmark | DeepSeek-V3 | Grok-3 mini |
|---|---|---|
| Fiction.LiveBench | 50% | 66.7% |
| LMArena Longer Query | 1353 | 1372 |

## Writing & Preference

- DeepSeek-V3: 57.6 (#124)
- Grok-3 mini: 52.7 (#163)

| Benchmark | DeepSeek-V3 | Grok-3 mini |
|---|---|---|
| LMArena Text | 1375 | 1370 |
| LMArena Creative Writing | 1364 | 1342 |
| Short-Story Creative Writing | 77% | 73.5% |
| WildBench | 83% | 65.1% |
| LMArena Multi-Turn | 1389 | 1355 |
| EQ-Bench Creative Writing | 1472 | — |
| LiveBench Language | 49.1% | — |

## FAQ

### Is DeepSeek-V3 better than Grok-3 mini?

Grok-3 mini is the stronger model overall, scoring 41.0 to 39.2 on the Noometry Index.

### Is DeepSeek-V3 or Grok-3 mini better for coding?

DeepSeek-V3 scores higher on coding benchmarks: 41.9 versus 40.4 in the Noometry coding category.

### How many benchmarks do DeepSeek-V3 and Grok-3 mini share?

33 benchmarks have published results for both models. DeepSeek-V3 has 60 scored results on Noometry and Grok-3 mini has 35.