# Claude Sonnet 4 vs Grok 4.6

> Grok 4.6 is the stronger model overall, scoring 56.8 to 40.6 on the Noometry Index.

- Canonical page: https://noometry.com/compare/claude-sonnet-4-vs-grok-4-6
- Last updated: 2026-10-10
- Shared benchmarks: 30

## Summary

- They share 30 benchmarks with published results for both. Claude Sonnet 4 scores higher in 0 categories and Grok 4.6 in 10 categories; 10 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Grok 4.6 leads 61.3 to 23.0.
- The biggest single-benchmark swing is ARC-AGI-2: 5.9% for Claude Sonnet 4 and 67.1% for Grok 4.6.
- Grok 4.6 is cheaper at $2 / $6 per million input/output tokens, against $3 / $15 for Claude Sonnet 4.
- Grok 4.6 accepts more context: 500K tokens versus 200K.

## Snapshot

| | Claude Sonnet 4 | Grok 4.6 |
|---|---|---|
| Provider | Anthropic | xAI |
| Noometry Index | 40.6 | 56.8 |
| Rank | 139 | 19 |
| Context | 200K | 500K |
| Input $/M | $3 | $2 |
| Output $/M | $15 | $6 |
| Weights | Proprietary | Proprietary |

## Coding

- Claude Sonnet 4: 44.0 (#70)
- Grok 4.6: 58.0 (#14)

| Benchmark | Claude Sonnet 4 | Grok 4.6 |
|---|---|---|
| SciCode | 40% | 56.5% |
| WeirdML | 46.1% | 67.3% |
| LMArena Coding | 1414 | 1464 |
| ALE-Bench | 655.35 | 1,508 |
| DeepSWE | — | 67.5% |
| FrontierCode | — | 48% |
| SWE-bench Verified (bash only) | 64.9% | — |
| Aider Polyglot | 61.3% | — |
| CursorBench | — | 41.4% |
| LMArena WebDev | — | 1620 |
| FrontierSWE | — | 25.3% |
| GSO | 4.9% | — |

## Agentic & Tool Use

- Claude Sonnet 4: 38.3 (#29)
- Grok 4.6: 39.4 (#25)

| Benchmark | Claude Sonnet 4 | Grok 4.6 |
|---|---|---|
| APEX-Agents | — | 65.3% |
| TheAgentCompany | 33.1% | — |
| Cybench | 35% | — |
| DeepResearch Bench | 46.6% | — |
| OSWorld | 43.9% | — |
| GDP.pdf | — | 17.2% |
| METR Time Horizons | 62% | — |
| Vending-Bench 2 | — | 9,047 |

## Reasoning

- Claude Sonnet 4: 23.0 (#169)
- Grok 4.6: 61.3 (#18)

| Benchmark | Claude Sonnet 4 | Grok 4.6 |
|---|---|---|
| ARC-AGI-2 | 5.9% | 67.1% |
| SimpleBench | 45.5% | 75.9% |
| ARC-AGI-1 | 40% | 87.5% |
| CritPt | 0.3% | 19.7% |
| LMArena Hard Prompts | 1372 | 1445 |
| DTBench | 77.1% | 97.3% |
| LMCA | 29% | 48.5% |
| Epoch Capabilities Index | 141.69 | 156.48 |
| Kagi LLM Benchmark | 73% | — |
| NYT Connections (extended) | — | 80% |
| Chess Puzzles | — | 40% |
| EnigmaEval | 3.1% | — |
| EBR-Bench | — | 30.5% |
| Mystery Game Puzzles | — | 34% |
| ForecastBench | 60.2 | — |

## Math

- Claude Sonnet 4: 41.5 (#71)
- Grok 4.6: 66.8 (#19)

| Benchmark | Claude Sonnet 4 | Grok 4.6 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 71.1% | 99.2% |
| LMArena Math | 1375 | 1416 |
| FrontierMath (Tiers 1-3) | — | 66% |
| FrontierMath Tier 4 | — | 31.7% |
| ProofBench | — | 51% |
| Omni-MATH | 60.2% | — |
| MATH Level 5 | 84.4% | — |
| FrontierMath (Feb 2025 set) | 4.1% | — |
| FrontierMath Tier 4 (v1) | 0% | — |

## Knowledge

- Claude Sonnet 4: 41.6 (#99)
- Grok 4.6: 63.3 (#18)

| Benchmark | Claude Sonnet 4 | Grok 4.6 |
|---|---|---|
| GPQA Diamond | 79.2% | 94% |
| LMArena Expert | 1371 | 1463 |
| Humanity's Last Exam | 7.8% | — |
| SimpleQA Verified | — | 49.3% |
| MMLU-Pro | 84.3% | — |
| Confabulations | 13.2% | — |
| Vectara Hallucination Rate | 10.3% | — |
| GPQA (HELM) | 70.6% | — |

## Multimodal

- Claude Sonnet 4: 25.6 (#116)
- Grok 4.6: 43.9 (#18)

| Benchmark | Claude Sonnet 4 | Grok 4.6 |
|---|---|---|
| LMArena Vision | 1191 | 1265 |
| GeoBench | 37% | — |
| VPCT | 34% | — |
| Blueprint-Bench 2 | — | 33.2% |
| Furniture Assembly | — | 40% |
| LMArena Document | — | 1452 |
| MindCube | 44.8% | — |

## Multilingual

- Claude Sonnet 4: 46.9 (#151)
- Grok 4.6: 53.3 (#67)

| Benchmark | Claude Sonnet 4 | Grok 4.6 |
|---|---|---|
| LMArena Non-English | 1333 | 1422 |
| LMArena Chinese | 1351 | 1479 |
| LMArena French | 1364 | 1458 |
| LMArena German | 1331 | 1425 |
| LMArena Japanese | 1301 | 1383 |
| LMArena Korean | 1293 | 1395 |
| LMArena Russian | 1355 | 1427 |
| LMArena Spanish | 1356 | 1408 |

## Instruction Following

- Claude Sonnet 4: 71.8 (#141)
- Grok 4.6: 75.6 (#59)

| Benchmark | Claude Sonnet 4 | Grok 4.6 |
|---|---|---|
| LMArena Instruction Following | 1375 | 1431 |
| IFEval | 84% | — |

## Long Context

- Claude Sonnet 4: 33.6 (#254)
- Grok 4.6: 44.4 (#62)

| Benchmark | Claude Sonnet 4 | Grok 4.6 |
|---|---|---|
| LMArena Longer Query | 1398 | 1453 |
| Fiction.LiveBench | 46.9% | — |

## Writing & Preference

- Claude Sonnet 4: 57.3 (#127)
- Grok 4.6: 62.5 (#77)

| Benchmark | Claude Sonnet 4 | Grok 4.6 |
|---|---|---|
| LMArena Text | 1351 | 1427 |
| LMArena Creative Writing | 1345 | 1430 |
| LMArena Multi-Turn | 1376 | 1422 |
| Short-Story Creative Writing | 81.4% | — |
| EQ-Bench Creative Writing | 1483 | — |
| WildBench | 83.8% | — |

## FAQ

### Is Claude Sonnet 4 better than Grok 4.6?

Grok 4.6 is the stronger model overall, scoring 56.8 to 40.6 on the Noometry Index.

### Which is cheaper, Claude Sonnet 4 or Grok 4.6?

Grok 4.6 is cheaper. It lists at $2 per million input tokens and $6 per million output tokens; Claude Sonnet 4 lists at $3 and $15.

### Is Claude Sonnet 4 or Grok 4.6 better for coding?

Grok 4.6 scores higher on coding benchmarks: 58.0 versus 44.0 in the Noometry coding category.

### Which has the bigger context window?

Grok 4.6 does, with 500K tokens against 200K.

### How many benchmarks do Claude Sonnet 4 and Grok 4.6 share?

30 benchmarks have published results for both models. Claude Sonnet 4 has 58 scored results on Noometry and Grok 4.6 has 49.