# GPT-6 Sol vs Grok 4

> GPT-6 Sol is the stronger model overall, scoring 61.8 to 48.1 on the Noometry Index.

- Canonical page: https://noometry.com/compare/gpt-6-sol-vs-grok-4
- Last updated: 2026-10-11
- Shared benchmarks: 23

## Summary

- They share 23 benchmarks with published results for both. GPT-6 Sol scores higher in 7 categories and Grok 4 in 3 categories; 10 gaps are clear of the uncertainty.
- The widest gap is in math, where GPT-6 Sol leads 87.2 to 48.4.
- The biggest single-benchmark swing is ARC-AGI-2: 89.6% for GPT-6 Sol and 16% for Grok 4.

## Snapshot

| | GPT-6 Sol | Grok 4 |
|---|---|---|
| Provider | OpenAI | xAI |
| Noometry Index | 61.8 | 48.1 |
| Rank | 12 | 56 |
| Context | 1.05M | — |
| Input $/M | $2 | — |
| Output $/M | $10 | — |
| Weights | Proprietary | Proprietary |

## Coding

- GPT-6 Sol: 60.1 (#11)
- Grok 4: 50.3 (#46)

| Benchmark | GPT-6 Sol | Grok 4 |
|---|---|---|
| LMArena Coding | 1447 | 1408 |
| DeepSWE | 68.8% | — |
| FrontierCode | 49.3% | — |
| Aider Polyglot | — | 79.6% |
| LMArena WebDev | 1688 | — |
| SciCode | 57.6% | — |
| WeirdML | — | 45.7% |
| ALE-Bench | 2,462 | — |

## Agentic & Tool Use

- GPT-6 Sol: 37.2 (#36)
- Grok 4: 32.3 (#68)

| Benchmark | GPT-6 Sol | Grok 4 |
|---|---|---|
| Terminal-Bench | — | 27.2% |
| APEX-Agents | 54.3% | — |
| Berkeley Function Calling Leaderboard | — | 63% |
| GDPval | — | 21.1% |
| Cybench | — | 43% |
| DeepResearch Bench | — | 47.3% |
| BALROG | — | 43.6% |
| GDP.pdf | 26.4% | — |
| LMArena Search | — | 1142 |
| METR Time Horizons | — | 66.6% |
| Vending-Bench 2 | 14,428 | — |

## Reasoning

- GPT-6 Sol: 74.0 (#9)
- Grok 4: 36.7 (#65)

| Benchmark | GPT-6 Sol | Grok 4 |
|---|---|---|
| ARC-AGI-2 | 89.6% | 16% |
| ARC-AGI-1 | 95.5% | 66.7% |
| LMArena Hard Prompts | 1418 | 1409 |
| Epoch Capabilities Index | 162.72 | 146.44 |
| SimpleBench | — | 60.5% |
| Kagi LLM Benchmark | — | 73.6% |
| NYT Connections (extended) | 90.1% | — |
| CritPt | 30.9% | — |
| Chess Puzzles | — | 28% |
| EBR-Bench | 53.3% | — |
| Mystery Game Puzzles | 56% | — |
| DTBench | 97.3% | — |
| LMCA | 59.1% | — |
| ForecastBench | — | 60.9 |

## Math

- GPT-6 Sol: 87.2 (#7)
- Grok 4: 48.4 (#64)

| Benchmark | GPT-6 Sol | Grok 4 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 100% | 84% |
| LMArena Math | 1402 | 1422 |
| FrontierMath (Tiers 1-3) | 89.8% | — |
| FrontierMath Tier 4 | 90% | — |
| ProofBench | 83% | — |
| Omni-MATH | — | 60.3% |
| FrontierMath (Feb 2025 set) | — | 19.7% |
| FrontierMath Tier 4 (v1) | — | 2.1% |

## Knowledge

- GPT-6 Sol: 64.8 (#15)
- Grok 4: 53.8 (#55)

| Benchmark | GPT-6 Sol | Grok 4 |
|---|---|---|
| GPQA Diamond | 94.3% | 87% |
| LMArena Expert | 1439 | 1415 |
| SimpleQA Verified | 60.7% | — |
| MMLU-Pro | — | 85.1% |
| Confabulations | — | 12.4% |
| Vectara Hallucination Rate | 6.5% | — |
| GPQA (HELM) | — | 72.7% |

## Multimodal

- GPT-6 Sol: 47.6 (#10)
- Grok 4: 33.7 (#94)

| Benchmark | GPT-6 Sol | Grok 4 |
|---|---|---|
| LMArena Vision | 1245 | 1210 |
| GeoBench | — | 45% |
| Blueprint-Bench 2 | 36.9% | — |
| Furniture Assembly | 58.3% | — |

## Multilingual

- GPT-6 Sol: 50.5 (#118)
- Grok 4: 51.8 (#103)

| Benchmark | GPT-6 Sol | Grok 4 |
|---|---|---|
| LMArena Non-English | 1385 | 1403 |
| LMArena Chinese | 1405 | 1427 |
| LMArena French | 1410 | 1418 |
| LMArena German | 1390 | 1429 |
| LMArena Japanese | 1385 | 1394 |
| LMArena Korean | 1341 | 1377 |
| LMArena Russian | 1401 | 1410 |
| LMArena Spanish | 1384 | 1420 |

## Instruction Following

- GPT-6 Sol: 74.5 (#94)
- Grok 4: 79.2 (#5)

| Benchmark | GPT-6 Sol | Grok 4 |
|---|---|---|
| LMArena Instruction Following | 1412 | 1387 |
| IFEval | — | 94.9% |

## Long Context

- GPT-6 Sol: 43.1 (#108)
- Grok 4: 63.1 (#4)

| Benchmark | GPT-6 Sol | Grok 4 |
|---|---|---|
| LMArena Longer Query | 1411 | 1409 |
| Fiction.LiveBench | — | 94.4% |

## Writing & Preference

- GPT-6 Sol: 71.9 (#18)
- Grok 4: 58.5 (#116)

| Benchmark | GPT-6 Sol | Grok 4 |
|---|---|---|
| LMArena Text | 1395 | 1411 |
| LMArena Creative Writing | 1378 | 1397 |
| LMArena Multi-Turn | 1412 | 1416 |
| Short-Story Creative Writing | — | 76.9% |
| EQ-Bench Creative Writing | 2125 | — |
| WildBench | — | 79.7% |

## FAQ

### Is GPT-6 Sol better than Grok 4?

GPT-6 Sol is the stronger model overall, scoring 61.8 to 48.1 on the Noometry Index.

### Is GPT-6 Sol or Grok 4 better for coding?

GPT-6 Sol scores higher on coding benchmarks: 60.1 versus 50.3 in the Noometry coding category.

### How many benchmarks do GPT-6 Sol and Grok 4 share?

23 benchmarks have published results for both models. GPT-6 Sol has 45 scored results on Noometry and Grok 4 has 48.