# Grok-2 (Dec 2024) vs Llama 3-70B

> Grok-2 (Dec 2024) is the stronger model overall, scoring 33.7 to 28.8 on the Noometry Index.

- Canonical page: https://noometry.com/compare/grok-2-vs-llama-3-70b
- Last updated: 2026-10-11
- Shared benchmarks: 22

## Summary

- They share 22 benchmarks with published results for both. Grok-2 (Dec 2024) scores higher in 6 categories and Llama 3-70B in 2 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in multilingual, where Grok-2 (Dec 2024) leads 43.1 to 33.6.
- The biggest single-benchmark swing is MATH Level 5: 63.5% for Grok-2 (Dec 2024) and 22.6% for Llama 3-70B.
- Llama 3-70B has downloadable open weights; the other is API-only.

## Snapshot

| | Grok-2 (Dec 2024) | Llama 3-70B |
|---|---|---|
| Provider | xAI | Meta |
| Noometry Index | 33.7 | 28.8 |
| Rank | 239 | 323 |
| Context | — | — |
| Input $/M | — | — |
| Output $/M | — | — |
| Weights | Proprietary | Open |

## Coding

- Grok-2 (Dec 2024): 33.3 (#258)
- Llama 3-70B: 35.8 (#218)

| Benchmark | Grok-2 (Dec 2024) | Llama 3-70B |
|---|---|---|
| LMArena Coding | 1287 | 1206 |
| WeirdML | 22.2% | — |
| BigCodeBench Instruct | — | 43.6% |
| LiveBench Coding | 46.4% | — |
| BigCodeBench Complete | — | 54.5% |
| HumanEval+ | — | 72% |
| MBPP+ | — | 69% |

## Agentic & Tool Use

- Grok-2 (Dec 2024): —
- Llama 3-70B: 21.1 (#139)

| Benchmark | Grok-2 (Dec 2024) | Llama 3-70B |
|---|---|---|
| Cybench | — | 5% |

## Reasoning

- Grok-2 (Dec 2024): 16.9 (#299)
- Llama 3-70B: 18.0 (#288)

| Benchmark | Grok-2 (Dec 2024) | Llama 3-70B |
|---|---|---|
| LMArena Hard Prompts | 1272 | 1195 |
| DTBench | 65.2% | 54.2% |
| Epoch Capabilities Index | 130.48 | 122.93 |
| SimpleBench | 22.7% | — |
| Kagi LLM Benchmark | — | 35.1% |
| LiveBench Reasoning | 54.8% | — |
| LiveBench Data Analysis | 54.5% | — |
| ForecastBench | — | 57.1 |
| LiveBench | 54.3% | — |
| WinoGrande | — | 83.5% |

## Math

- Grok-2 (Dec 2024): 20.8 (#284)
- Llama 3-70B: 12.8 (#305)

| Benchmark | Grok-2 (Dec 2024) | Llama 3-70B |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 11.5% | 4.3% |
| LMArena Math | 1283 | 1218 |
| MATH Level 5 | 63.5% | 22.6% |
| LiveBench Math | 54.9% | — |
| FrontierMath (Feb 2025 set) | 0.7% | — |

## Knowledge

- Grok-2 (Dec 2024): 29.8 (#233)
- Llama 3-70B: 20.8 (#277)

| Benchmark | Grok-2 (Dec 2024) | Llama 3-70B |
|---|---|---|
| GPQA Diamond | 53.8% | 40.6% |
| LMArena Expert | 1254 | 1149 |
| Confabulations | 20.1% | — |
| MMLU | — | 79.3% |

## Multilingual

- Grok-2 (Dec 2024): 43.1 (#188)
- Llama 3-70B: 33.6 (#251)

| Benchmark | Grok-2 (Dec 2024) | Llama 3-70B |
|---|---|---|
| LMArena Non-English | 1282 | 1142 |
| LMArena Chinese | 1289 | 1114 |
| LMArena French | 1318 | 1232 |
| LMArena German | 1287 | 1169 |
| LMArena Japanese | 1244 | 1017 |
| LMArena Korean | 1237 | 1017 |
| LMArena Russian | 1286 | 1159 |
| LMArena Spanish | 1281 | 1241 |

## Instruction Following

- Grok-2 (Dec 2024): 66.9 (#202)
- Llama 3-70B: 62.5 (#238)

| Benchmark | Grok-2 (Dec 2024) | Llama 3-70B |
|---|---|---|
| LMArena Instruction Following | 1270 | 1194 |
| LiveBench Instruction Following | 69.6% | — |

## Long Context

- Grok-2 (Dec 2024): 38.8 (#190)
- Llama 3-70B: 35.6 (#240)

| Benchmark | Grok-2 (Dec 2024) | Llama 3-70B |
|---|---|---|
| LMArena Longer Query | 1276 | 1174 |

## Writing & Preference

- Grok-2 (Dec 2024): 48.6 (#198)
- Llama 3-70B: 42.8 (#231)

| Benchmark | Grok-2 (Dec 2024) | Llama 3-70B |
|---|---|---|
| LMArena Text | 1305 | 1221 |
| LMArena Creative Writing | 1284 | 1210 |
| LMArena Multi-Turn | 1290 | 1223 |
| Short-Story Creative Writing | 63.6% | — |
| LiveBench Language | 45.6% | — |

## FAQ

### Is Grok-2 (Dec 2024) better than Llama 3-70B?

Grok-2 (Dec 2024) is the stronger model overall, scoring 33.7 to 28.8 on the Noometry Index.

### Is Grok-2 (Dec 2024) or Llama 3-70B better for coding?

Llama 3-70B scores higher on coding benchmarks: 35.8 versus 33.3 in the Noometry coding category.

### How many benchmarks do Grok-2 (Dec 2024) and Llama 3-70B share?

22 benchmarks have published results for both models. Grok-2 (Dec 2024) has 34 scored results on Noometry and Llama 3-70B has 31.