# GPT-4.5 vs Grok 4

> Grok 4 is the stronger model overall, scoring 48.1 to 37.2 on the Noometry Index.

- Canonical page: https://noometry.com/compare/gpt-4-5-vs-grok-4
- Last updated: 2026-10-11
- Shared benchmarks: 30

## Summary

- They share 30 benchmarks with published results for both. GPT-4.5 scores higher in 2 categories and Grok 4 in 8 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Grok 4 leads 36.7 to 13.9.
- The biggest single-benchmark swing is ARC-AGI-1: 10.3% for GPT-4.5 and 66.7% for Grok 4.

## Snapshot

| | GPT-4.5 | Grok 4 |
|---|---|---|
| Provider | OpenAI | xAI |
| Noometry Index | 37.2 | 48.1 |
| Rank | 208 | 56 |
| Context | — | — |
| Input $/M | — | — |
| Output $/M | — | — |
| Weights | Proprietary | Proprietary |

## Coding

- GPT-4.5: 42.2 (#109)
- Grok 4: 50.3 (#46)

| Benchmark | GPT-4.5 | Grok 4 |
|---|---|---|
| Aider Polyglot | 44.9% | 79.6% |
| WeirdML | 39.4% | 45.7% |
| LMArena Coding | 1396 | 1408 |
| LiveBench Coding | 75.2% | — |

## Agentic & Tool Use

- GPT-4.5: 27.9 (#97)
- Grok 4: 32.3 (#68)

| Benchmark | GPT-4.5 | Grok 4 |
|---|---|---|
| Cybench | 17.5% | 43% |
| Terminal-Bench | — | 27.2% |
| Berkeley Function Calling Leaderboard | — | 63% |
| GDPval | — | 21.1% |
| DeepResearch Bench | — | 47.3% |
| BALROG | — | 43.6% |
| LMArena Search | — | 1142 |
| METR Time Horizons | — | 66.6% |

## Reasoning

- GPT-4.5: 13.9 (#330)
- Grok 4: 36.7 (#65)

| Benchmark | GPT-4.5 | Grok 4 |
|---|---|---|
| ARC-AGI-2 | 0.8% | 16% |
| SimpleBench | 34.5% | 60.5% |
| ARC-AGI-1 | 10.3% | 66.7% |
| LMArena Hard Prompts | 1403 | 1409 |
| Epoch Capabilities Index | 136.74 | 146.44 |
| ForecastBench | 61.7 | 60.9 |
| Kagi LLM Benchmark | — | 73.6% |
| Chess Puzzles | — | 28% |
| EnigmaEval | 3.2% | — |
| LiveBench Reasoning | 71.1% | — |
| LiveBench Data Analysis | 64.3% | — |
| LiveBench | 69% | — |

## Math

- GPT-4.5: 32.6 (#211)
- Grok 4: 48.4 (#64)

| Benchmark | GPT-4.5 | Grok 4 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 37.8% | 84% |
| LMArena Math | 1412 | 1422 |
| Omni-MATH | — | 60.3% |
| LiveBench Math | 69.3% | — |
| MATH Level 5 | 78.6% | — |
| FrontierMath (Feb 2025 set) | — | 19.7% |
| FrontierMath Tier 4 (v1) | — | 2.1% |

## Knowledge

- GPT-4.5: 32.5 (#211)
- Grok 4: 53.8 (#55)

| Benchmark | GPT-4.5 | Grok 4 |
|---|---|---|
| GPQA Diamond | 68.7% | 87% |
| Confabulations | 13.6% | 12.4% |
| LMArena Expert | 1394 | 1415 |
| Humanity's Last Exam | 5.4% | — |
| MMLU-Pro | — | 85.1% |
| GPQA (HELM) | — | 72.7% |

## Multimodal

- GPT-4.5: 37.6 (#71)
- Grok 4: 33.7 (#94)

| Benchmark | GPT-4.5 | Grok 4 |
|---|---|---|
| LMArena Vision | 1195 | 1210 |
| GeoBench | — | 45% |
| VPCT | 45% | — |

## Multilingual

- GPT-4.5: 52.5 (#83)
- Grok 4: 51.8 (#103)

| Benchmark | GPT-4.5 | Grok 4 |
|---|---|---|
| LMArena Non-English | 1413 | 1403 |
| LMArena Chinese | 1421 | 1427 |
| LMArena French | 1418 | 1418 |
| LMArena German | 1457 | 1429 |
| LMArena Japanese | 1416 | 1394 |
| LMArena Korean | 1392 | 1377 |
| LMArena Russian | 1419 | 1410 |
| LMArena Spanish | — | 1420 |

## Instruction Following

- GPT-4.5: 72.6 (#134)
- Grok 4: 79.2 (#5)

| Benchmark | GPT-4.5 | Grok 4 |
|---|---|---|
| LMArena Instruction Following | 1404 | 1387 |
| LiveBench Instruction Following | 72.3% | — |
| IFEval | — | 94.9% |

## Long Context

- GPT-4.5: 40.4 (#155)
- Grok 4: 63.1 (#4)

| Benchmark | GPT-4.5 | Grok 4 |
|---|---|---|
| Fiction.LiveBench | 63.9% | 94.4% |
| LMArena Longer Query | 1406 | 1409 |

## Writing & Preference

- GPT-4.5: 56.9 (#134)
- Grok 4: 58.5 (#116)

| Benchmark | GPT-4.5 | Grok 4 |
|---|---|---|
| LMArena Text | 1417 | 1411 |
| LMArena Creative Writing | 1394 | 1397 |
| Short-Story Creative Writing | 75.6% | 76.9% |
| LMArena Multi-Turn | 1444 | 1416 |
| EQ-Bench Creative Writing | 1258 | — |
| WildBench | — | 79.7% |
| LiveBench Language | 61.5% | — |

## FAQ

### Is GPT-4.5 better than Grok 4?

Grok 4 is the stronger model overall, scoring 48.1 to 37.2 on the Noometry Index.

### Is GPT-4.5 or Grok 4 better for coding?

Grok 4 scores higher on coding benchmarks: 50.3 versus 42.2 in the Noometry coding category.

### How many benchmarks do GPT-4.5 and Grok 4 share?

30 benchmarks have published results for both models. GPT-4.5 has 42 scored results on Noometry and Grok 4 has 48.