Noometry

Model comparison

Grok 4 vs Llama-3.3-70B-Instruct

Grok 4 is the stronger model overall, scoring 48.1 to 30.6 on the Noometry Index.

Last verified . 27 shared benchmarks.

Grok 4 xAI

48.1

Rank #56 Confirmed

Llama-3.3-70B-Instruct Meta

30.6

Rank #291 Confirmed

Summary

  • They share 27 benchmarks with published results for both. Grok 4 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 leads 63.1 to 26.4.
  • The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 84% for Grok 4 and 5.1% for Llama-3.3-70B-Instruct.
  • Llama-3.3-70B-Instruct has downloadable open weights; the other is API-only.

Side by side

Grok 4Llama-3.3-70B-Instruct
ProviderxAIMeta
Noometry Index48.130.6
Released2025-07-092024-12-06
WeightsProprietaryOpen
Context window—128K
Max output—4K
Input $ / M tokens—$0.10
Output $ / M tokens—$0.32
Results tracked4843

Sponsored placements are available on pages like this one. Advertise on Noometry

Category by category

Coding Grok 4 leads

Grok 4: 50.3 (#46), Llama-3.3-70B-Instruct: 31.0 (#290)

BenchmarkGrok 4Llama-3.3-70B-Instruct
WeirdML45.7%14.4%
LMArena Coding14081268
Aider Polyglot79.6%—
SciCode—26%
BigCodeBench Instruct—46.9%
LiveBench Coding—36.6%
BigCodeBench Complete—57.5%

Agentic & Tool Use Grok 4 leads

Grok 4: 32.3 (#68), Llama-3.3-70B-Instruct: 25.8 (#105)

BenchmarkGrok 4Llama-3.3-70B-Instruct
Berkeley Function Calling Leaderboard63%31.9%
BALROG43.6%23%
Terminal-Bench27.2%—
GDPval21.1%—
Cybench43%—
DeepResearch Bench47.3%—
LMArena Search1142—
METR Time Horizons66.6%—

Reasoning Grok 4 leads

Grok 4: 36.7 (#65), Llama-3.3-70B-Instruct: 14.1 (#327)

BenchmarkGrok 4Llama-3.3-70B-Instruct
SimpleBench60.5%19.9%
LMArena Hard Prompts14091257
Epoch Capabilities Index146.44127.33
ForecastBench60.958.6
ARC-AGI-216%—
Kagi LLM Benchmark73.6%—
ARC-AGI-166.7%—
CritPt—0%
Chess Puzzles28%—
LiveBench Reasoning—50.8%
DTBench—59.5%
LiveBench Data Analysis—49.5%
LMCA—17.5%
LiveBench—50.2%

Math Grok 4 leads

Grok 4: 48.4 (#64), Llama-3.3-70B-Instruct: 15.3 (#298)

BenchmarkGrok 4Llama-3.3-70B-Instruct
OTIS Mock AIME 2024-202584%5.1%
LMArena Math14221267
Omni-MATH60.3%—
LiveBench Math—42.2%
MATH Level 5—41.6%
FrontierMath (Feb 2025 set)19.7%—
FrontierMath Tier 4 (v1)2.1%—

Knowledge Grok 4 leads

Grok 4: 53.8 (#55), Llama-3.3-70B-Instruct: 30.6 (#226)

BenchmarkGrok 4Llama-3.3-70B-Instruct
GPQA Diamond87%47.4%
Confabulations12.4%22.8%
LMArena Expert14151225
MMLU-Pro85.1%—
Vectara Hallucination Rate—4.1%
GPQA (HELM)72.7%—
MMLU—86.3%

Multimodal Not comparable

Grok 4: 33.7 (#94), Llama-3.3-70B-Instruct: —

BenchmarkGrok 4Llama-3.3-70B-Instruct
LMArena Vision1210—
GeoBench45%—

Multilingual Grok 4 leads

Grok 4: 51.8 (#103), Llama-3.3-70B-Instruct: 39.9 (#220)

BenchmarkGrok 4Llama-3.3-70B-Instruct
LMArena Non-English14031236
LMArena Chinese14271217
LMArena French14181281
LMArena German14291251
LMArena Japanese13941150
LMArena Korean13771143
LMArena Russian14101252
LMArena Spanish14201270

Instruction Following Grok 4 leads

Grok 4: 79.2 (#5), Llama-3.3-70B-Instruct: 71.1 (#157)

BenchmarkGrok 4Llama-3.3-70B-Instruct
LMArena Instruction Following13871242
LiveBench Instruction Following—82.7%
IFEval94.9%—

Long Context Grok 4 leads

Grok 4: 63.1 (#4), Llama-3.3-70B-Instruct: 26.4 (#295)

BenchmarkGrok 4Llama-3.3-70B-Instruct
Fiction.LiveBench94.4%33.3%
LMArena Longer Query14091256

Writing & Preference Grok 4 leads

Grok 4: 58.5 (#116), Llama-3.3-70B-Instruct: 47.6 (#207)

BenchmarkGrok 4Llama-3.3-70B-Instruct
LMArena Text14111274
LMArena Creative Writing13971250
LMArena Multi-Turn14161280
Short-Story Creative Writing76.9%—
WildBench79.7%—
LiveBench Language—39.2%

Frequently asked questions

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

Grok 4 is the stronger model overall, scoring 48.1 to 30.6 on the Noometry Index.

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

Grok 4 scores higher on coding benchmarks: 50.3 versus 31.0 in the Noometry coding category.

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

27 benchmarks have published results for both models. Grok 4 has 48 scored results on Noometry and Llama-3.3-70B-Instruct has 43.

Related comparisons

Go deeper