# GPT-4.5 vs Grok-2 (Dec 2024)

> GPT-4.5 is the stronger model overall, scoring 36.9 to 33.4 on the Noometry Index.

- Canonical page: https://noometry.com/compare/gpt-4-5-vs-grok-2
- Last updated: 2026-10-10
- Shared benchmarks: 31

## Summary

- They share 31 benchmarks with published results for both. GPT-4.5 scores higher in 7 categories and Grok-2 (Dec 2024) in 1 category; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where GPT-4.5 leads 31.0 to 19.6.
- The biggest single-benchmark swing is LiveBench Coding: 75.2% for GPT-4.5 and 46.4% for Grok-2 (Dec 2024).

## Snapshot

| | GPT-4.5 | Grok-2 (Dec 2024) |
|---|---|---|
| Provider | OpenAI | xAI |
| Noometry Index | 36.9 | 33.4 |
| Rank | 200 | 233 |
| Context | — | — |
| Input $/M | — | — |
| Output $/M | — | — |
| Weights | Proprietary | Proprietary |

## Coding

- GPT-4.5: 41.8 (#102)
- Grok-2 (Dec 2024): 32.9 (#254)

| Benchmark | GPT-4.5 | Grok-2 (Dec 2024) |
|---|---|---|
| WeirdML | 39.4% | 22.2% |
| LiveBench Coding | 75.2% | 46.4% |
| LMArena Coding | 1397 | 1288 |
| Aider Polyglot | 44.9% | — |

## Agentic & Tool Use

- GPT-4.5: 27.6 (#92)
- Grok-2 (Dec 2024): —

| Benchmark | GPT-4.5 | Grok-2 (Dec 2024) |
|---|---|---|
| Cybench | 17.5% | — |

## Reasoning

- GPT-4.5: 14.1 (#320)
- Grok-2 (Dec 2024): 16.4 (#294)

| Benchmark | GPT-4.5 | Grok-2 (Dec 2024) |
|---|---|---|
| SimpleBench | 34.5% | 22.7% |
| LiveBench Reasoning | 71.1% | 54.8% |
| LMArena Hard Prompts | 1404 | 1272 |
| LiveBench Data Analysis | 64.3% | 54.5% |
| Epoch Capabilities Index | 136.74 | 130.48 |
| LiveBench | 69% | 54.3% |
| ARC-AGI-2 | 0.8% | — |
| ARC-AGI-1 | 10.3% | — |
| EnigmaEval | 3.2% | — |
| DTBench | — | 65.2% |
| ForecastBench | 61.7 | — |

## Math

- GPT-4.5: 31.0 (#200)
- Grok-2 (Dec 2024): 19.6 (#278)

| Benchmark | GPT-4.5 | Grok-2 (Dec 2024) |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 37.8% | 11.5% |
| LiveBench Math | 69.3% | 54.9% |
| LMArena Math | 1412 | 1283 |
| MATH Level 5 | 78.6% | 63.5% |
| FrontierMath (Feb 2025 set) | — | 0.7% |

## Knowledge

- GPT-4.5: 32.4 (#205)
- Grok-2 (Dec 2024): 29.8 (#225)

| Benchmark | GPT-4.5 | Grok-2 (Dec 2024) |
|---|---|---|
| GPQA Diamond | 68.7% | 53.8% |
| Confabulations | 13.6% | 20.1% |
| LMArena Expert | 1394 | 1254 |
| Humanity's Last Exam | 5.4% | — |

## Multimodal

- GPT-4.5: 37.0 (#68)
- Grok-2 (Dec 2024): —

| Benchmark | GPT-4.5 | Grok-2 (Dec 2024) |
|---|---|---|
| LMArena Vision | 1195 | — |
| VPCT | 45% | — |

## Multilingual

- GPT-4.5: 52.7 (#78)
- Grok-2 (Dec 2024): 43.3 (#183)

| Benchmark | GPT-4.5 | Grok-2 (Dec 2024) |
|---|---|---|
| LMArena Non-English | 1413 | 1282 |
| LMArena Chinese | 1421 | 1289 |
| LMArena French | 1417 | 1317 |
| LMArena German | 1457 | 1287 |
| LMArena Japanese | 1416 | 1244 |
| LMArena Korean | 1392 | 1237 |
| LMArena Russian | 1419 | 1286 |
| LMArena Spanish | — | 1281 |

## Instruction Following

- GPT-4.5: 72.8 (#129)
- Grok-2 (Dec 2024): 67.1 (#197)

| Benchmark | GPT-4.5 | Grok-2 (Dec 2024) |
|---|---|---|
| LiveBench Instruction Following | 72.3% | 69.6% |
| LMArena Instruction Following | 1404 | 1270 |

## Long Context

- GPT-4.5: 40.4 (#150)
- Grok-2 (Dec 2024): 38.7 (#186)

| Benchmark | GPT-4.5 | Grok-2 (Dec 2024) |
|---|---|---|
| LMArena Longer Query | 1406 | 1276 |
| Fiction.LiveBench | 63.9% | — |

## Writing & Preference

- GPT-4.5: 57.1 (#129)
- Grok-2 (Dec 2024): 48.8 (#192)

| Benchmark | GPT-4.5 | Grok-2 (Dec 2024) |
|---|---|---|
| LMArena Text | 1417 | 1305 |
| LMArena Creative Writing | 1394 | 1284 |
| Short-Story Creative Writing | 75.6% | 63.6% |
| LMArena Multi-Turn | 1444 | 1290 |
| LiveBench Language | 61.5% | 45.6% |
| EQ-Bench Creative Writing | 1258 | — |

## FAQ

### Is GPT-4.5 better than Grok-2 (Dec 2024)?

GPT-4.5 is the stronger model overall, scoring 36.9 to 33.4 on the Noometry Index.

### Is GPT-4.5 or Grok-2 (Dec 2024) better for coding?

GPT-4.5 scores higher on coding benchmarks: 41.8 versus 32.9 in the Noometry coding category.

### How many benchmarks do GPT-4.5 and Grok-2 (Dec 2024) share?

31 benchmarks have published results for both models. GPT-4.5 has 42 scored results on Noometry and Grok-2 (Dec 2024) has 34.