# DeepSeek-V3.2-Exp vs Grok 3

> DeepSeek-V3.2-Exp is the stronger model overall, scoring 44.3 to 39.9 on the Noometry Index.

- Canonical page: https://noometry.com/compare/deepseek-v3-2-exp-vs-grok-3
- Last updated: 2026-10-10
- Shared benchmarks: 30

## Summary

- They share 30 benchmarks with published results for both. DeepSeek-V3.2-Exp scores higher in 7 categories and Grok 3 in 2 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in long context, where DeepSeek-V3.2-Exp leads 47.6 to 38.7.
- The biggest single-benchmark swing is ARC-AGI-1: 57% for DeepSeek-V3.2-Exp and 5.5% for Grok 3.
- DeepSeek-V3.2-Exp has downloadable open weights; the other is API-only.

## Snapshot

| | DeepSeek-V3.2-Exp | Grok 3 |
|---|---|---|
| Provider | DeepSeek | xAI |
| Noometry Index | 44.3 | 39.9 |
| Rank | 78 | 157 |
| Context | 164K | — |
| Input $/M | $0.26 | — |
| Output $/M | $0.38 | — |
| Weights | Open | Proprietary |

## Coding

- DeepSeek-V3.2-Exp: 46.5 (#65)
- Grok 3: 41.9 (#115)

| Benchmark | DeepSeek-V3.2-Exp | Grok 3 |
|---|---|---|
| Aider Polyglot | 74.2% | 53.3% |
| WeirdML | 39.5% | 37.2% |
| LMArena Coding | 1454 | 1432 |
| SWE-bench Verified (bash only) | 70% | — |
| LMArena WebDev | 1362 | — |
| SWE-bench Multilingual | 59% | — |
| SciCode | 38.9% | — |

## Agentic & Tool Use

- DeepSeek-V3.2-Exp: 32.7 (#59)
- Grok 3: 30.5 (#76)

| Benchmark | DeepSeek-V3.2-Exp | Grok 3 |
|---|---|---|
| Terminal-Bench | 39.6% | — |
| APEX-Agents | 21.3% | — |
| Berkeley Function Calling Leaderboard | 56.7% | — |
| TheAgentCompany | 42.9% | — |
| BALROG | — | 29.5% |
| Vending-Bench 2 | 1,034 | — |

## Reasoning

- DeepSeek-V3.2-Exp: 22.1 (#208)
- Grok 3: 13.7 (#333)

| Benchmark | DeepSeek-V3.2-Exp | Grok 3 |
|---|---|---|
| ARC-AGI-2 | 4% | 0% |
| Kagi LLM Benchmark | 52.2% | 61.3% |
| ARC-AGI-1 | 57% | 5.5% |
| LMArena Hard Prompts | 1434 | 1434 |
| Epoch Capabilities Index | 146.27 | 138.33 |
| SimpleBench | — | 36.1% |
| NYT Connections (extended) | 36.7% | — |
| CritPt | 2.9% | — |
| Chess Puzzles | 14% | — |
| Thematic Generalization | 65% | — |
| DTBench | 87.7% | — |
| LMCA | 29.1% | — |

## Math

- DeepSeek-V3.2-Exp: 41.7 (#87)
- Grok 3: 38.0 (#145)

| Benchmark | DeepSeek-V3.2-Exp | Grok 3 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 87.8% | 55.6% |
| LMArena Math | 1435 | 1391 |
| FrontierMath (Feb 2025 set) | 22.1% | 3.8% |
| FrontierMath Tier 4 (v1) | 2.1% | 0% |
| MathArena Final-Answer Competitions | 57.7% | — |
| ProofBench | 8% | — |
| Omni-MATH | — | 46.4% |
| MATH Level 5 | — | 88.7% |

## Knowledge

- DeepSeek-V3.2-Exp: 51.7 (#66)
- Grok 3: 46.2 (#82)

| Benchmark | DeepSeek-V3.2-Exp | Grok 3 |
|---|---|---|
| GPQA Diamond | 83.4% | 75.8% |
| Vectara Hallucination Rate | 5.3% | 5.8% |
| LMArena Expert | 1436 | 1421 |
| MMLU-Pro | — | 78.8% |
| Confabulations | — | 14.2% |
| GPQA (HELM) | — | 65% |

## Multilingual

- DeepSeek-V3.2-Exp: 52.2 (#90)
- Grok 3: 52.3 (#87)

| Benchmark | DeepSeek-V3.2-Exp | Grok 3 |
|---|---|---|
| LMArena Non-English | 1409 | 1410 |
| LMArena Chinese | 1461 | 1448 |
| LMArena French | 1433 | 1460 |
| LMArena German | 1440 | 1431 |
| LMArena Japanese | 1374 | 1387 |
| LMArena Korean | 1371 | 1373 |
| LMArena Russian | 1424 | 1416 |
| LMArena Spanish | 1440 | 1417 |

## Instruction Following

- DeepSeek-V3.2-Exp: 74.5 (#93)
- Grok 3: 75.0 (#73)

| Benchmark | DeepSeek-V3.2-Exp | Grok 3 |
|---|---|---|
| LMArena Instruction Following | 1413 | 1409 |
| IFEval | — | 88.4% |

## Long Context

- DeepSeek-V3.2-Exp: 47.6 (#16)
- Grok 3: 38.7 (#192)

| Benchmark | DeepSeek-V3.2-Exp | Grok 3 |
|---|---|---|
| Fiction.LiveBench | 83.3% | 58.3% |
| LMArena Longer Query | 1428 | 1439 |
| CL-bench | 13.2% | — |
| CL-bench Life | 9.5% | — |

## Writing & Preference

- DeepSeek-V3.2-Exp: 62.4 (#77)
- Grok 3: 55.8 (#141)

| Benchmark | DeepSeek-V3.2-Exp | Grok 3 |
|---|---|---|
| LMArena Text | 1425 | 1426 |
| LMArena Creative Writing | 1403 | 1414 |
| EQ-Bench Creative Writing | 1515 | 1186 |
| LMArena Multi-Turn | 1427 | 1425 |
| Short-Story Creative Writing | — | 76.4% |
| WildBench | — | 84.9% |

## FAQ

### Is DeepSeek-V3.2-Exp better than Grok 3?

DeepSeek-V3.2-Exp is the stronger model overall, scoring 44.3 to 39.9 on the Noometry Index.

### Is DeepSeek-V3.2-Exp or Grok 3 better for coding?

DeepSeek-V3.2-Exp scores higher on coding benchmarks: 46.5 versus 41.9 in the Noometry coding category.

### How many benchmarks do DeepSeek-V3.2-Exp and Grok 3 share?

30 benchmarks have published results for both models. DeepSeek-V3.2-Exp has 49 scored results on Noometry and Grok 3 has 40.