# Grok 3 vs Llama-3.3-70B-Instruct

> Grok 3 is the stronger model overall, scoring 39.9 to 30.6 on the Noometry Index.

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

## Summary

- They share 27 benchmarks with published results for both. Grok 3 scores higher in 8 categories and Llama-3.3-70B-Instruct in 1 category; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where Grok 3 leads 38.0 to 15.3.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 55.6% for Grok 3 and 5.1% for Llama-3.3-70B-Instruct.
- Llama-3.3-70B-Instruct has downloadable open weights; the other is API-only.

## Snapshot

| | Grok 3 | Llama-3.3-70B-Instruct |
|---|---|---|
| Provider | xAI | Meta |
| Noometry Index | 39.9 | 30.6 |
| Rank | 157 | 291 |
| Context | — | 128K |
| Input $/M | — | $0.10 |
| Output $/M | — | $0.32 |
| Weights | Proprietary | Open |

## Coding

- Grok 3: 41.9 (#115)
- Llama-3.3-70B-Instruct: 31.0 (#290)

| Benchmark | Grok 3 | Llama-3.3-70B-Instruct |
|---|---|---|
| WeirdML | 37.2% | 14.4% |
| LMArena Coding | 1432 | 1268 |
| Aider Polyglot | 53.3% | — |
| SciCode | — | 26% |
| BigCodeBench Instruct | — | 46.9% |
| LiveBench Coding | — | 36.6% |
| BigCodeBench Complete | — | 57.5% |

## Agentic & Tool Use

- Grok 3: 30.5 (#76)
- Llama-3.3-70B-Instruct: 25.8 (#105)

| Benchmark | Grok 3 | Llama-3.3-70B-Instruct |
|---|---|---|
| BALROG | 29.5% | 23% |
| Berkeley Function Calling Leaderboard | — | 31.9% |

## Reasoning

- Grok 3: 13.7 (#333)
- Llama-3.3-70B-Instruct: 14.1 (#327)

| Benchmark | Grok 3 | Llama-3.3-70B-Instruct |
|---|---|---|
| SimpleBench | 36.1% | 19.9% |
| LMArena Hard Prompts | 1434 | 1257 |
| Epoch Capabilities Index | 138.33 | 127.33 |
| ARC-AGI-2 | 0% | — |
| Kagi LLM Benchmark | 61.3% | — |
| ARC-AGI-1 | 5.5% | — |
| CritPt | — | 0% |
| LiveBench Reasoning | — | 50.8% |
| DTBench | — | 59.5% |
| LiveBench Data Analysis | — | 49.5% |
| LMCA | — | 17.5% |
| ForecastBench | — | 58.6 |
| LiveBench | — | 50.2% |

## Math

- Grok 3: 38.0 (#145)
- Llama-3.3-70B-Instruct: 15.3 (#298)

| Benchmark | Grok 3 | Llama-3.3-70B-Instruct |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 55.6% | 5.1% |
| LMArena Math | 1391 | 1267 |
| MATH Level 5 | 88.7% | 41.6% |
| Omni-MATH | 46.4% | — |
| LiveBench Math | — | 42.2% |
| FrontierMath (Feb 2025 set) | 3.8% | — |
| FrontierMath Tier 4 (v1) | 0% | — |

## Knowledge

- Grok 3: 46.2 (#82)
- Llama-3.3-70B-Instruct: 30.6 (#226)

| Benchmark | Grok 3 | Llama-3.3-70B-Instruct |
|---|---|---|
| GPQA Diamond | 75.8% | 47.4% |
| Confabulations | 14.2% | 22.8% |
| Vectara Hallucination Rate | 5.8% | 4.1% |
| LMArena Expert | 1421 | 1225 |
| MMLU-Pro | 78.8% | — |
| GPQA (HELM) | 65% | — |
| MMLU | — | 86.3% |

## Multilingual

- Grok 3: 52.3 (#87)
- Llama-3.3-70B-Instruct: 39.9 (#220)

| Benchmark | Grok 3 | Llama-3.3-70B-Instruct |
|---|---|---|
| LMArena Non-English | 1410 | 1236 |
| LMArena Chinese | 1448 | 1217 |
| LMArena French | 1460 | 1281 |
| LMArena German | 1431 | 1251 |
| LMArena Japanese | 1387 | 1150 |
| LMArena Korean | 1373 | 1143 |
| LMArena Russian | 1416 | 1252 |
| LMArena Spanish | 1417 | 1270 |

## Instruction Following

- Grok 3: 75.0 (#73)
- Llama-3.3-70B-Instruct: 71.1 (#157)

| Benchmark | Grok 3 | Llama-3.3-70B-Instruct |
|---|---|---|
| LMArena Instruction Following | 1409 | 1242 |
| LiveBench Instruction Following | — | 82.7% |
| IFEval | 88.4% | — |

## Long Context

- Grok 3: 38.7 (#192)
- Llama-3.3-70B-Instruct: 26.4 (#295)

| Benchmark | Grok 3 | Llama-3.3-70B-Instruct |
|---|---|---|
| Fiction.LiveBench | 58.3% | 33.3% |
| LMArena Longer Query | 1439 | 1256 |

## Writing & Preference

- Grok 3: 55.8 (#141)
- Llama-3.3-70B-Instruct: 47.6 (#207)

| Benchmark | Grok 3 | Llama-3.3-70B-Instruct |
|---|---|---|
| LMArena Text | 1426 | 1274 |
| LMArena Creative Writing | 1414 | 1250 |
| LMArena Multi-Turn | 1425 | 1280 |
| Short-Story Creative Writing | 76.4% | — |
| EQ-Bench Creative Writing | 1186 | — |
| WildBench | 84.9% | — |
| LiveBench Language | — | 39.2% |

## FAQ

### Is Grok 3 better than Llama-3.3-70B-Instruct?

Grok 3 is the stronger model overall, scoring 39.9 to 30.6 on the Noometry Index.

### Is Grok 3 or Llama-3.3-70B-Instruct better for coding?

Grok 3 scores higher on coding benchmarks: 41.9 versus 31.0 in the Noometry coding category.

### How many benchmarks do Grok 3 and Llama-3.3-70B-Instruct share?

27 benchmarks have published results for both models. Grok 3 has 40 scored results on Noometry and Llama-3.3-70B-Instruct has 43.