# Grok 4 vs Llama 3.1-70B

> Grok 4 is the stronger model overall, scoring 47.8 to 29.4 on the Noometry Index.

- Canonical page: https://noometry.com/compare/grok-4-vs-llama-3-1-70b
- Last updated: 2026-10-10
- Shared benchmarks: 27

## Summary

- They share 27 benchmarks with published results for both. Grok 4 scores higher in 9 categories and Llama 3.1-70B in 0 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in math, where Grok 4 leads 46.5 to 12.7.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 84% for Grok 4 and 3.6% for Llama 3.1-70B.
- Llama 3.1-70B has downloadable open weights; the other is API-only.

## Snapshot

| | Grok 4 | Llama 3.1-70B |
|---|---|---|
| Provider | xAI | Meta |
| Noometry Index | 47.8 | 29.4 |
| Rank | 49 | 299 |
| Context | — | 128K |
| Input $/M | — | $0.40 |
| Output $/M | — | $0.40 |
| Weights | Proprietary | Open |

## Coding

- Grok 4: 49.9 (#44)
- Llama 3.1-70B: 29.8 (#292)

| Benchmark | Grok 4 | Llama 3.1-70B |
|---|---|---|
| WeirdML | 45.7% | 9% |
| LMArena Coding | 1408 | 1260 |
| Aider Polyglot | 79.6% | — |
| BigCodeBench Instruct | — | 46.1% |
| BigCodeBench Complete | — | 54.8% |

## Agentic & Tool Use

- Grok 4: 32.1 (#62)
- Llama 3.1-70B: 24.9 (#104)

| Benchmark | Grok 4 | Llama 3.1-70B |
|---|---|---|
| BALROG | 43.6% | 27.9% |
| Terminal-Bench | 27.2% | — |
| Berkeley Function Calling Leaderboard | 63% | — |
| GDPval | 21.1% | — |
| TheAgentCompany | — | 6.9% |
| Cybench | 43% | — |
| DeepResearch Bench | 47.3% | — |
| LMArena Search | 1142 | — |
| METR Time Horizons | 66.6% | — |

## Reasoning

- Grok 4: 36.9 (#59)
- Llama 3.1-70B: 20.8 (#215)

| Benchmark | Grok 4 | Llama 3.1-70B |
|---|---|---|
| LMArena Hard Prompts | 1409 | 1241 |
| Epoch Capabilities Index | 146.46 | 125.91 |
| ARC-AGI-2 | 16% | — |
| SimpleBench | 60.5% | — |
| Kagi LLM Benchmark | 73.6% | — |
| ARC-AGI-1 | 66.7% | — |
| Chess Puzzles | 28% | — |
| DTBench | — | 60% |
| LMCA | — | 14.8% |
| ForecastBench | 60.9 | — |

## Math

- Grok 4: 46.5 (#58)
- Llama 3.1-70B: 12.7 (#298)

| Benchmark | Grok 4 | Llama 3.1-70B |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 84% | 3.6% |
| Omni-MATH | 60.3% | 21% |
| LMArena Math | 1422 | 1252 |
| MATH Level 5 | — | 36.7% |
| FrontierMath (Feb 2025 set) | 19.7% | — |
| FrontierMath Tier 4 (v1) | 2.1% | — |

## Knowledge

- Grok 4: 53.7 (#50)
- Llama 3.1-70B: 24.1 (#264)

| Benchmark | Grok 4 | Llama 3.1-70B |
|---|---|---|
| GPQA Diamond | 87% | 44.2% |
| MMLU-Pro | 85.1% | 65.3% |
| GPQA (HELM) | 72.7% | 42.6% |
| LMArena Expert | 1417 | 1209 |
| Confabulations | 12.4% | — |
| MMLU | — | 80.1% |

## Multimodal

- Grok 4: 33.1 (#89)
- Llama 3.1-70B: —

| Benchmark | Grok 4 | Llama 3.1-70B |
|---|---|---|
| LMArena Vision | 1211 | — |
| GeoBench | 45% | — |

## Multilingual

- Grok 4: 52.0 (#98)
- Llama 3.1-70B: 38.9 (#220)

| Benchmark | Grok 4 | Llama 3.1-70B |
|---|---|---|
| LMArena Non-English | 1403 | 1219 |
| LMArena Chinese | 1428 | 1215 |
| LMArena French | 1418 | 1260 |
| LMArena German | 1429 | 1222 |
| LMArena Japanese | 1396 | 1132 |
| LMArena Korean | 1376 | 1140 |
| LMArena Russian | 1409 | 1234 |
| LMArena Spanish | 1419 | 1253 |

## Instruction Following

- Grok 4: 79.4 (#4)
- Llama 3.1-70B: 65.5 (#218)

| Benchmark | Grok 4 | Llama 3.1-70B |
|---|---|---|
| IFEval | 94.9% | 82.1% |
| LMArena Instruction Following | 1387 | 1231 |

## Long Context

- Grok 4: 63.1 (#4)
- Llama 3.1-70B: 37.6 (#208)

| Benchmark | Grok 4 | Llama 3.1-70B |
|---|---|---|
| LMArena Longer Query | 1409 | 1241 |
| Fiction.LiveBench | 94.4% | — |

## Writing & Preference

- Grok 4: 58.7 (#111)
- Llama 3.1-70B: 35.6 (#262)

| Benchmark | Grok 4 | Llama 3.1-70B |
|---|---|---|
| LMArena Text | 1412 | 1261 |
| LMArena Creative Writing | 1397 | 1232 |
| WildBench | 79.7% | 75.8% |
| LMArena Multi-Turn | 1416 | 1256 |
| Short-Story Creative Writing | 76.9% | — |
| EQ-Bench Creative Writing | — | 784 |

## FAQ

### Is Grok 4 better than Llama 3.1-70B?

Grok 4 is the stronger model overall, scoring 47.8 to 29.4 on the Noometry Index.

### Is Grok 4 or Llama 3.1-70B better for coding?

Grok 4 scores higher on coding benchmarks: 49.9 versus 29.8 in the Noometry coding category.

### How many benchmarks do Grok 4 and Llama 3.1-70B share?

27 benchmarks have published results for both models. Grok 4 has 48 scored results on Noometry and Llama 3.1-70B has 35.