Noometry

Model comparison

Grok 4 vs Llama 3.1-70B

Grok 4 is the stronger model overall, scoring 48.1 to 29.6 on the Noometry Index.

Last verified . 27 shared benchmarks.

Grok 4 xAI

48.1

Rank #56 Confirmed

Llama 3.1-70B Meta

29.6

Rank #308 Confirmed

Summary

  • They share 27 benchmarks with published results for both. Grok 4 scores higher in 9 categories and Llama 3.1-70B in 0 categories; 9 gaps are clear of the uncertainty.
  • The widest gap is in math, where Grok 4 leads 48.4 to 13.5.
  • The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 84% for Grok 4 and 3.6% for Llama 3.1-70B.
  • Llama 3.1-70B has downloadable open weights; the other is API-only.

Side by side

Grok 4Llama 3.1-70B
ProviderxAIMeta
Noometry Index48.129.6
Released2025-07-092024-07-23
WeightsProprietaryOpen
Context window—128K
Max output—4K
Input $ / M tokens—$0.40
Output $ / M tokens—$0.40
Results tracked4835

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 3.1-70B: 30.3 (#296)

BenchmarkGrok 4Llama 3.1-70B
WeirdML45.7%9%
LMArena Coding14081260
Aider Polyglot79.6%—
BigCodeBench Instruct—46.1%
BigCodeBench Complete—54.8%

Agentic & Tool Use Grok 4 leads

Grok 4: 32.3 (#68), Llama 3.1-70B: 25.1 (#112)

BenchmarkGrok 4Llama 3.1-70B
BALROG43.6%27.9%
Terminal-Bench27.2%—
Berkeley Function Calling Leaderboard63%—
GDPval21.1%—
TheAgentCompany—6.9%
Cybench43%—
DeepResearch Bench47.3%—
LMArena Search1142—
METR Time Horizons66.6%—

Reasoning Grok 4 leads

Grok 4: 36.7 (#65), Llama 3.1-70B: 21.6 (#220)

BenchmarkGrok 4Llama 3.1-70B
LMArena Hard Prompts14091241
Epoch Capabilities Index146.44125.92
ARC-AGI-216%—
SimpleBench60.5%—
Kagi LLM Benchmark73.6%—
ARC-AGI-166.7%—
Chess Puzzles28%—
DTBench—60%
LMCA—14.8%
ForecastBench60.9—

Math Grok 4 leads

Grok 4: 48.4 (#64), Llama 3.1-70B: 13.5 (#304)

BenchmarkGrok 4Llama 3.1-70B
OTIS Mock AIME 2024-202584%3.6%
Omni-MATH60.3%21%
LMArena Math14221252
MATH Level 5—36.7%
FrontierMath (Feb 2025 set)19.7%—
FrontierMath Tier 4 (v1)2.1%—

Knowledge Grok 4 leads

Grok 4: 53.8 (#55), Llama 3.1-70B: 24.2 (#269)

BenchmarkGrok 4Llama 3.1-70B
GPQA Diamond87%44.2%
MMLU-Pro85.1%65.3%
GPQA (HELM)72.7%42.6%
LMArena Expert14151209
Confabulations12.4%—
MMLU—80.1%

Multimodal Not comparable

Grok 4: 33.7 (#94), Llama 3.1-70B: —

BenchmarkGrok 4Llama 3.1-70B
LMArena Vision1210—
GeoBench45%—

Multilingual Grok 4 leads

Grok 4: 51.8 (#103), Llama 3.1-70B: 38.8 (#225)

BenchmarkGrok 4Llama 3.1-70B
LMArena Non-English14031219
LMArena Chinese14271215
LMArena French14181261
LMArena German14291222
LMArena Japanese13941132
LMArena Korean13771140
LMArena Russian14101234
LMArena Spanish14201253

Instruction Following Grok 4 leads

Grok 4: 79.2 (#5), Llama 3.1-70B: 65.3 (#223)

BenchmarkGrok 4Llama 3.1-70B
IFEval94.9%82.1%
LMArena Instruction Following13871231

Long Context Grok 4 leads

Grok 4: 63.1 (#4), Llama 3.1-70B: 37.6 (#214)

BenchmarkGrok 4Llama 3.1-70B
LMArena Longer Query14091241
Fiction.LiveBench94.4%—

Writing & Preference Grok 4 leads

Grok 4: 58.5 (#116), Llama 3.1-70B: 35.4 (#267)

BenchmarkGrok 4Llama 3.1-70B
LMArena Text14111261
LMArena Creative Writing13971232
WildBench79.7%75.8%
LMArena Multi-Turn14161256
Short-Story Creative Writing76.9%—
EQ-Bench Creative Writing—784

Frequently asked questions

Is Grok 4 better than Llama 3.1-70B?

Grok 4 is the stronger model overall, scoring 48.1 to 29.6 on the Noometry Index.

Is Grok 4 or Llama 3.1-70B better for coding?

Grok 4 scores higher on coding benchmarks: 50.3 versus 30.3 in the Noometry coding category.

How many benchmarks do Grok 4 and Llama 3.1-70B share?

27 benchmarks have published results for both models. Grok 4 has 48 scored results on Noometry and Llama 3.1-70B has 35.

Related comparisons

Go deeper