# DeepSeek-V3.1 vs Grok 4.5

> Grok 4.5 is the stronger model overall, scoring 55.0 to 42.2 on the Noometry Index. DeepSeek-V3.1 costs 7.1× less per token, which makes it the better buy when Grok 4.5's lead doesn't matter for your workload.

- Canonical page: https://noometry.com/compare/deepseek-v3-1-vs-grok-4-5
- Last updated: 2026-10-10
- Shared benchmarks: 24

## Summary

- They share 24 benchmarks with published results for both. DeepSeek-V3.1 scores higher in 0 categories and Grok 4.5 in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Grok 4.5 leads 56.0 to 27.3.
- The biggest single-benchmark swing is Kagi LLM Benchmark: 53.2% for DeepSeek-V3.1 and 83.5% for Grok 4.5.
- DeepSeek-V3.1 is cheaper at $0.25 / $0.95 per million input/output tokens, against $2 / $6 for Grok 4.5.
- Grok 4.5 accepts more context: 500K tokens versus 164K.
- DeepSeek-V3.1 has downloadable open weights; the other is API-only.

## Snapshot

| | DeepSeek-V3.1 | Grok 4.5 |
|---|---|---|
| Provider | DeepSeek | xAI |
| Noometry Index | 42.2 | 55.0 |
| Rank | 106 | 22 |
| Context | 164K | 500K |
| Input $/M | $0.25 | $2 |
| Output $/M | $0.95 | $6 |
| Weights | Open | Proprietary |

## Coding

- DeepSeek-V3.1: 39.8 (#140)
- Grok 4.5: 51.7 (#34)

| Benchmark | DeepSeek-V3.1 | Grok 4.5 |
|---|---|---|
| WeirdML | 38.4% | 46.4% |
| LMArena Coding | 1417 | 1473 |
| DeepSWE | — | 53.8% |
| FrontierCode | — | 42.4% |
| LMArena WebDev | — | 1552 |
| SciCode | — | 54.1% |
| ALE-Bench | — | 1,309 |

## Agentic & Tool Use

- DeepSeek-V3.1: —
- Grok 4.5: 44.4 (#15)

| Benchmark | DeepSeek-V3.1 | Grok 4.5 |
|---|---|---|
| APEX-Agents | — | 56.2% |
| τ²-bench Banking | — | 47.9% |
| PostTrainBench | — | 23.4% |
| GBAEval | — | 65.4% |
| GDP.pdf | — | 14% |
| LMArena Search | — | 1213 |
| Vending-Bench 2 | — | 3,887 |

## Reasoning

- DeepSeek-V3.1: 27.3 (#109)
- Grok 4.5: 56.0 (#23)

| Benchmark | DeepSeek-V3.1 | Grok 4.5 |
|---|---|---|
| SimpleBench | 40% | 70% |
| Kagi LLM Benchmark | 53.2% | 83.5% |
| LMArena Hard Prompts | 1417 | 1462 |
| DTBench | 82.7% | 96.5% |
| LMCA | 24.3% | 45.2% |
| Epoch Capabilities Index | 139.92 | 154.02 |
| ARC-AGI-2 | — | 52.6% |
| NYT Connections (extended) | — | 79.9% |
| ARC-AGI-1 | — | 87.2% |
| CritPt | — | 15.4% |
| Chess Puzzles | — | 36% |
| Surface Evolver Bench | — | 74.4% |
| ForecastBench | 58 | — |

## Math

- DeepSeek-V3.1: 36.1 (#125)
- Grok 4.5: 60.7 (#29)

| Benchmark | DeepSeek-V3.1 | Grok 4.5 |
|---|---|---|
| LMArena Math | 1420 | 1457 |
| FrontierMath (Tiers 1-3) | — | 57.2% |
| FrontierMath Tier 4 | — | 24.4% |
| OTIS Mock AIME 2024-2025 | — | 97.8% |
| ProofBench | — | 31% |

## Knowledge

- DeepSeek-V3.1: 43.0 (#86)
- Grok 4.5: 62.3 (#21)

| Benchmark | DeepSeek-V3.1 | Grok 4.5 |
|---|---|---|
| LMArena Expert | 1403 | 1466 |
| GPQA Diamond | — | 93.4% |
| SimpleQA Verified | — | 48.3% |
| Vectara Hallucination Rate | 5.5% | — |

## Multimodal

- DeepSeek-V3.1: —
- Grok 4.5: 37.8 (#63)

| Benchmark | DeepSeek-V3.1 | Grok 4.5 |
|---|---|---|
| LMArena Vision | — | 1287 |
| Blueprint-Bench 2 | — | 27.3% |
| Furniture Assembly | — | 22.5% |
| LMArena Document | — | 1452 |

## Multilingual

- DeepSeek-V3.1: 51.8 (#100)
- Grok 4.5: 54.5 (#43)

| Benchmark | DeepSeek-V3.1 | Grok 4.5 |
|---|---|---|
| LMArena Non-English | 1400 | 1439 |
| LMArena Chinese | 1470 | 1497 |
| LMArena French | 1447 | 1458 |
| LMArena German | 1412 | 1448 |
| LMArena Japanese | 1374 | 1425 |
| LMArena Korean | 1337 | 1402 |
| LMArena Russian | 1405 | 1447 |
| LMArena Spanish | 1429 | 1452 |

## Instruction Following

- DeepSeek-V3.1: 74.1 (#105)
- Grok 4.5: 76.3 (#45)

| Benchmark | DeepSeek-V3.1 | Grok 4.5 |
|---|---|---|
| LMArena Instruction Following | 1400 | 1446 |

## Long Context

- DeepSeek-V3.1: 36.3 (#227)
- Grok 4.5: 44.8 (#54)

| Benchmark | DeepSeek-V3.1 | Grok 4.5 |
|---|---|---|
| LMArena Longer Query | 1422 | 1463 |
| Fiction.LiveBench | 52.8% | — |

## Writing & Preference

- DeepSeek-V3.1: 60.5 (#93)
- Grok 4.5: 65.9 (#40)

| Benchmark | DeepSeek-V3.1 | Grok 4.5 |
|---|---|---|
| LMArena Text | 1420 | 1448 |
| LMArena Creative Writing | 1401 | 1442 |
| EQ-Bench Creative Writing | 1436 | 1579 |
| LMArena Multi-Turn | 1408 | 1457 |

## FAQ

### Is DeepSeek-V3.1 better than Grok 4.5?

Grok 4.5 is the stronger model overall, scoring 55.0 to 42.2 on the Noometry Index. DeepSeek-V3.1 costs 7.1× less per token, which makes it the better buy when Grok 4.5's lead doesn't matter for your workload.

### Which is cheaper, DeepSeek-V3.1 or Grok 4.5?

DeepSeek-V3.1 is cheaper. It lists at $0.25 per million input tokens and $0.95 per million output tokens; Grok 4.5 lists at $2 and $6.

### Is DeepSeek-V3.1 or Grok 4.5 better for coding?

Grok 4.5 scores higher on coding benchmarks: 51.7 versus 39.8 in the Noometry coding category.

### Which has the bigger context window?

Grok 4.5 does, with 500K tokens against 164K.

### How many benchmarks do DeepSeek-V3.1 and Grok 4.5 share?

24 benchmarks have published results for both models. DeepSeek-V3.1 has 27 scored results on Noometry and Grok 4.5 has 52.