# Grok-3 mini vs Llama 2-70B

> Grok-3 mini is the stronger model overall, scoring 41.2 to 24.4 on the Noometry Index.

- Canonical page: https://noometry.com/compare/grok-3-mini-vs-llama-2-70b
- Last updated: 2026-10-10
- Shared benchmarks: 21

## Summary

- They share 21 benchmarks with published results for both. Grok-3 mini scores higher in 7 categories and Llama 2-70B in 1 category; 7 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where Grok-3 mini leads 46.4 to 7.4.
- The biggest single-benchmark swing is MATH Level 5: 90.9% for Grok-3 mini and 3.3% for Llama 2-70B.
- Llama 2-70B has downloadable open weights; the other is API-only.

## Snapshot

| | Grok-3 mini | Llama 2-70B |
|---|---|---|
| Provider | xAI | Meta |
| Noometry Index | 41.2 | 24.4 |
| Rank | 141 | 349 |
| Context | — | — |
| Input $/M | — | — |
| Output $/M | — | — |
| Weights | Proprietary | Open |

## Coding

- Grok-3 mini: 40.8 (#131)
- Llama 2-70B: 31.4 (#286)

| Benchmark | Grok-3 mini | Llama 2-70B |
|---|---|---|
| LMArena Coding | 1379 | 1079 |
| Aider Polyglot | 49.3% | — |
| WeirdML | 42.6% | — |

## Reasoning

- Grok-3 mini: 13.6 (#334)
- Llama 2-70B: 14.4 (#325)

| Benchmark | Grok-3 mini | Llama 2-70B |
|---|---|---|
| LMArena Hard Prompts | 1375 | 1073 |
| Epoch Capabilities Index | 140.35 | 113.79 |
| ARC-AGI-2 | 0.4% | — |
| Kagi LLM Benchmark | 61.3% | — |
| ARC-AGI-1 | 16.5% | — |
| DTBench | — | 41.6% |
| BIG-Bench Hard | — | 64.9% |
| CommonsenseQA 2.0 | — | 50% |
| ForecastBench | — | 51.4 |
| HellaSwag | — | 85.3% |
| LAMBADA | — | 78.9% |
| PIQA | — | 82.8% |
| WinoGrande | — | 80.2% |

## Math

- Grok-3 mini: 42.1 (#85)
- Llama 2-70B: 8.1 (#326)

| Benchmark | Grok-3 mini | Llama 2-70B |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 77.8% | 0% |
| LMArena Math | 1386 | 1091 |
| MATH Level 5 | 90.9% | 3.3% |
| Omni-MATH | 31.8% | — |
| FrontierMath (Feb 2025 set) | 5.9% | — |
| GSM8K | — | 69.6% |

## Knowledge

- Grok-3 mini: 46.4 (#81)
- Llama 2-70B: 7.4 (#310)

| Benchmark | Grok-3 mini | Llama 2-70B |
|---|---|---|
| GPQA Diamond | 76.3% | 26.3% |
| LMArena Expert | 1395 | 1039 |
| MMLU-Pro | 79.9% | — |
| Confabulations | 10.8% | — |
| GPQA (HELM) | 67.5% | — |
| ARC (AI2) Challenge | — | 78.3% |
| BoolQ | — | 88.6% |
| MMLU | — | 69.9% |
| OpenBookQA | — | 60.2% |
| TriviaQA | — | 87.6% |

## Multilingual

- Grok-3 mini: 48.1 (#145)
- Llama 2-70B: 27.7 (#274)

| Benchmark | Grok-3 mini | Llama 2-70B |
|---|---|---|
| LMArena Non-English | 1352 | 1045 |
| LMArena Chinese | 1387 | 995 |
| LMArena French | 1357 | 1090 |
| LMArena German | 1349 | 1041 |
| LMArena Japanese | 1342 | 927 |
| LMArena Korean | 1335 | 964 |
| LMArena Russian | 1353 | 1083 |
| LMArena Spanish | 1381 | 1143 |

## Instruction Following

- Grok-3 mini: 78.5 (#9)
- Llama 2-70B: 54.9 (#278)

| Benchmark | Grok-3 mini | Llama 2-70B |
|---|---|---|
| LMArena Instruction Following | 1357 | 1071 |
| IFEval | 95.1% | — |

## Long Context

- Grok-3 mini: 41.0 (#147)
- Llama 2-70B: 32.3 (#270)

| Benchmark | Grok-3 mini | Llama 2-70B |
|---|---|---|
| LMArena Longer Query | 1372 | 1062 |
| Fiction.LiveBench | 66.7% | — |

## Writing & Preference

- Grok-3 mini: 52.5 (#169)
- Llama 2-70B: 32.3 (#279)

| Benchmark | Grok-3 mini | Llama 2-70B |
|---|---|---|
| LMArena Text | 1370 | 1115 |
| LMArena Creative Writing | 1342 | 1075 |
| LMArena Multi-Turn | 1355 | 1088 |
| Short-Story Creative Writing | 73.5% | — |
| WildBench | 65.1% | — |

## FAQ

### Is Grok-3 mini better than Llama 2-70B?

Grok-3 mini is the stronger model overall, scoring 41.2 to 24.4 on the Noometry Index.

### Is Grok-3 mini or Llama 2-70B better for coding?

Grok-3 mini scores higher on coding benchmarks: 40.8 versus 31.4 in the Noometry coding category.

### How many benchmarks do Grok-3 mini and Llama 2-70B share?

21 benchmarks have published results for both models. Grok-3 mini has 35 scored results on Noometry and Llama 2-70B has 35.