# GPT-5.4 nano vs Grok-2 (Dec 2024)

> GPT-5.4 nano is the stronger model overall, scoring 41.9 to 33.4 on the Noometry Index.

- Canonical page: https://noometry.com/compare/gpt-5-4-nano-vs-grok-2
- Last updated: 2026-10-10
- Shared benchmarks: 23

## Summary

- They share 23 benchmarks with published results for both. GPT-5.4 nano scores higher in 8 categories and Grok-2 (Dec 2024) in 0 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where GPT-5.4 nano leads 40.7 to 19.6.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 87.8% for GPT-5.4 nano and 11.5% for Grok-2 (Dec 2024).

## Snapshot

| | GPT-5.4 nano | Grok-2 (Dec 2024) |
|---|---|---|
| Provider | OpenAI | xAI |
| Noometry Index | 41.9 | 33.4 |
| Rank | 111 | 233 |
| Context | 400K | — |
| Input $/M | $0.20 | — |
| Output $/M | $1.25 | — |
| Weights | Proprietary | Proprietary |

## Coding

- GPT-5.4 nano: 43.1 (#82)
- Grok-2 (Dec 2024): 32.9 (#254)

| Benchmark | GPT-5.4 nano | Grok-2 (Dec 2024) |
|---|---|---|
| WeirdML | 49.2% | 22.2% |
| LMArena Coding | 1404 | 1288 |
| SciCode | 46.9% | — |
| LiveBench Coding | — | 46.4% |
| ALE-Bench | 1,005 | — |

## Reasoning

- GPT-5.4 nano: 23.8 (#159)
- Grok-2 (Dec 2024): 16.4 (#294)

| Benchmark | GPT-5.4 nano | Grok-2 (Dec 2024) |
|---|---|---|
| LMArena Hard Prompts | 1380 | 1272 |
| DTBench | 80.3% | 65.2% |
| Epoch Capabilities Index | 145.78 | 130.48 |
| ARC-AGI-2 | 5.7% | — |
| SimpleBench | — | 22.7% |
| Kagi LLM Benchmark | 39.7% | — |
| ARC-AGI-1 | 51.5% | — |
| CritPt | 9.3% | — |
| Chess Puzzles | 30% | — |
| LiveBench Reasoning | — | 54.8% |
| Mystery Game Puzzles | 9% | — |
| LiveBench Data Analysis | — | 54.5% |
| LMCA | 36.9% | — |
| ForecastBench | 57.3 | — |
| LiveBench | — | 54.3% |

## Math

- GPT-5.4 nano: 40.7 (#76)
- Grok-2 (Dec 2024): 19.6 (#278)

| Benchmark | GPT-5.4 nano | Grok-2 (Dec 2024) |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 87.8% | 11.5% |
| LMArena Math | 1406 | 1283 |
| FrontierMath (Feb 2025 set) | 25.9% | 0.7% |
| FrontierMath (Tiers 1-3) | 44.9% | — |
| FrontierMath Tier 4 | 12.2% | — |
| ProofBench | 5% | — |
| LiveBench Math | — | 54.9% |
| MATH Level 5 | — | 63.5% |
| FrontierMath Tier 4 (v1) | 6.3% | — |

## Knowledge

- GPT-5.4 nano: 41.8 (#94)
- Grok-2 (Dec 2024): 29.8 (#225)

| Benchmark | GPT-5.4 nano | Grok-2 (Dec 2024) |
|---|---|---|
| GPQA Diamond | 78.5% | 53.8% |
| LMArena Expert | 1397 | 1254 |
| SimpleQA Verified | 11.7% | — |
| Confabulations | — | 20.1% |
| Vectara Hallucination Rate | 3.1% | — |

## Multimodal

- GPT-5.4 nano: 36.0 (#73)
- Grok-2 (Dec 2024): —

| Benchmark | GPT-5.4 nano | Grok-2 (Dec 2024) |
|---|---|---|
| LMArena Vision | 1196 | — |

## Multilingual

- GPT-5.4 nano: 48.8 (#136)
- Grok-2 (Dec 2024): 43.3 (#183)

| Benchmark | GPT-5.4 nano | Grok-2 (Dec 2024) |
|---|---|---|
| LMArena Non-English | 1359 | 1282 |
| LMArena Chinese | 1391 | 1289 |
| LMArena French | 1396 | 1317 |
| LMArena German | 1368 | 1287 |
| LMArena Japanese | 1342 | 1244 |
| LMArena Korean | 1321 | 1237 |
| LMArena Russian | 1363 | 1286 |
| LMArena Spanish | 1368 | 1281 |

## Instruction Following

- GPT-5.4 nano: 72.1 (#139)
- Grok-2 (Dec 2024): 67.1 (#197)

| Benchmark | GPT-5.4 nano | Grok-2 (Dec 2024) |
|---|---|---|
| LMArena Instruction Following | 1362 | 1270 |
| LiveBench Instruction Following | — | 69.6% |

## Long Context

- GPT-5.4 nano: 41.6 (#131)
- Grok-2 (Dec 2024): 38.7 (#186)

| Benchmark | GPT-5.4 nano | Grok-2 (Dec 2024) |
|---|---|---|
| LMArena Longer Query | 1365 | 1276 |

## Writing & Preference

- GPT-5.4 nano: 55.9 (#137)
- Grok-2 (Dec 2024): 48.8 (#192)

| Benchmark | GPT-5.4 nano | Grok-2 (Dec 2024) |
|---|---|---|
| LMArena Text | 1372 | 1305 |
| LMArena Creative Writing | 1313 | 1284 |
| LMArena Multi-Turn | 1382 | 1290 |
| Short-Story Creative Writing | — | 63.6% |
| LiveBench Language | — | 45.6% |

## FAQ

### Is GPT-5.4 nano better than Grok-2 (Dec 2024)?

GPT-5.4 nano is the stronger model overall, scoring 41.9 to 33.4 on the Noometry Index.

### Is GPT-5.4 nano or Grok-2 (Dec 2024) better for coding?

GPT-5.4 nano scores higher on coding benchmarks: 43.1 versus 32.9 in the Noometry coding category.

### How many benchmarks do GPT-5.4 nano and Grok-2 (Dec 2024) share?

23 benchmarks have published results for both models. GPT-5.4 nano has 40 scored results on Noometry and Grok-2 (Dec 2024) has 34.