# Grok 4 Fast vs Llama-3.3-70B-Instruct

> Grok 4 Fast is the stronger model overall, scoring 39.0 to 30.3 on the Noometry Index.

- Canonical page: https://noometry.com/compare/grok-4-fast-vs-llama-3-3-70b-instruct
- Last updated: 2026-10-10
- Shared benchmarks: 23

## Summary

- They share 23 benchmarks with published results for both. Grok 4 Fast scores higher in 9 categories and Llama-3.3-70B-Instruct in 0 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in long context, where Grok 4 Fast leads 63.2 to 26.4.
- The biggest single-benchmark swing is Fiction.LiveBench: 94.4% for Grok 4 Fast and 33.3% for Llama-3.3-70B-Instruct.
- Llama-3.3-70B-Instruct has downloadable open weights; the other is API-only.

## Snapshot

| | Grok 4 Fast | Llama-3.3-70B-Instruct |
|---|---|---|
| Provider | xAI | Meta |
| Noometry Index | 39.0 | 30.3 |
| Rank | 164 | 283 |
| Context | — | 128K |
| Input $/M | — | $0.10 |
| Output $/M | — | $0.32 |
| Weights | Proprietary | Open |

## Coding

- Grok 4 Fast: 32.3 (#264)
- Llama-3.3-70B-Instruct: 30.6 (#284)

| Benchmark | Grok 4 Fast | Llama-3.3-70B-Instruct |
|---|---|---|
| WeirdML | 42.9% | 14.4% |
| LMArena Coding | 1429 | 1268 |
| LMArena WebDev | 1160 | — |
| SciCode | — | 26% |
| BigCodeBench Instruct | — | 46.9% |
| LiveBench Coding | — | 36.6% |
| BigCodeBench Complete | — | 57.5% |

## Agentic & Tool Use

- Grok 4 Fast: 29.3 (#80)
- Llama-3.3-70B-Instruct: 25.6 (#100)

| Benchmark | Grok 4 Fast | Llama-3.3-70B-Instruct |
|---|---|---|
| Berkeley Function Calling Leaderboard | — | 31.9% |
| τ²-bench Banking | 15.7% | — |
| Cybench | 30% | — |
| BALROG | — | 23% |
| LMArena Search | 1171 | — |

## Reasoning

- Grok 4 Fast: 22.5 (#179)
- Llama-3.3-70B-Instruct: 13.7 (#325)

| Benchmark | Grok 4 Fast | Llama-3.3-70B-Instruct |
|---|---|---|
| LMArena Hard Prompts | 1413 | 1257 |
| DTBench | 82.7% | 59.5% |
| Epoch Capabilities Index | 144.21 | 127.32 |
| ForecastBench | 60.5 | 58.6 |
| ARC-AGI-2 | 5.3% | — |
| SimpleBench | — | 19.9% |
| Kagi LLM Benchmark | 66.1% | — |
| ARC-AGI-1 | 48.5% | — |
| CritPt | — | 0% |
| LiveBench Reasoning | — | 50.8% |
| LiveBench Data Analysis | — | 49.5% |
| LMCA | — | 17.5% |
| LiveBench | — | 50.2% |

## Math

- Grok 4 Fast: 36.1 (#126)
- Llama-3.3-70B-Instruct: 14.3 (#292)

| Benchmark | Grok 4 Fast | Llama-3.3-70B-Instruct |
|---|---|---|
| LMArena Math | 1419 | 1268 |
| OTIS Mock AIME 2024-2025 | — | 5.1% |
| LiveBench Math | — | 42.2% |
| MATH Level 5 | — | 41.6% |

## Knowledge

- Grok 4 Fast: 31.5 (#211)
- Llama-3.3-70B-Instruct: 30.5 (#220)

| Benchmark | Grok 4 Fast | Llama-3.3-70B-Instruct |
|---|---|---|
| Vectara Hallucination Rate | 19.7% | 4.1% |
| LMArena Expert | 1411 | 1225 |
| GPQA Diamond | — | 47.4% |
| Confabulations | — | 22.8% |
| MMLU | — | 86.3% |

## Multilingual

- Grok 4 Fast: 51.5 (#106)
- Llama-3.3-70B-Instruct: 40.1 (#215)

| Benchmark | Grok 4 Fast | Llama-3.3-70B-Instruct |
|---|---|---|
| LMArena Non-English | 1397 | 1236 |
| LMArena Chinese | 1458 | 1217 |
| LMArena French | 1433 | 1280 |
| LMArena German | 1383 | 1251 |
| LMArena Japanese | 1358 | 1149 |
| LMArena Korean | 1359 | 1142 |
| LMArena Russian | 1387 | 1252 |
| LMArena Spanish | 1414 | 1269 |

## Instruction Following

- Grok 4 Fast: 73.4 (#116)
- Llama-3.3-70B-Instruct: 71.3 (#152)

| Benchmark | Grok 4 Fast | Llama-3.3-70B-Instruct |
|---|---|---|
| LMArena Instruction Following | 1388 | 1242 |
| LiveBench Instruction Following | — | 82.7% |

## Long Context

- Grok 4 Fast: 63.2 (#3)
- Llama-3.3-70B-Instruct: 26.4 (#290)

| Benchmark | Grok 4 Fast | Llama-3.3-70B-Instruct |
|---|---|---|
| Fiction.LiveBench | 94.4% | 33.3% |
| LMArena Longer Query | 1415 | 1256 |

## Writing & Preference

- Grok 4 Fast: 60.2 (#96)
- Llama-3.3-70B-Instruct: 47.8 (#203)

| Benchmark | Grok 4 Fast | Llama-3.3-70B-Instruct |
|---|---|---|
| LMArena Text | 1407 | 1274 |
| LMArena Creative Writing | 1388 | 1250 |
| LMArena Multi-Turn | 1414 | 1279 |
| LiveBench Language | — | 39.2% |

## FAQ

### Is Grok 4 Fast better than Llama-3.3-70B-Instruct?

Grok 4 Fast is the stronger model overall, scoring 39.0 to 30.3 on the Noometry Index.

### Is Grok 4 Fast or Llama-3.3-70B-Instruct better for coding?

Grok 4 Fast scores higher on coding benchmarks: 32.3 versus 30.6 in the Noometry coding category.

### How many benchmarks do Grok 4 Fast and Llama-3.3-70B-Instruct share?

23 benchmarks have published results for both models. Grok 4 Fast has 30 scored results on Noometry and Llama-3.3-70B-Instruct has 43.