Noometry

Model comparison

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

Grok-3 mini is the stronger model overall, scoring 41.2 to 30.6 on the Noometry Index.

Last verified . 24 shared benchmarks.

Grok-3 mini xAI

41.2

Rank #141 Confirmed

Llama-3.3-70B-Instruct Meta

30.6

Rank #291 Confirmed

Summary

  • They share 24 benchmarks with published results for both. Grok-3 mini scores higher in 7 categories and Llama-3.3-70B-Instruct in 1 category; 7 gaps are clear of the uncertainty.
  • The widest gap is in math, where Grok-3 mini leads 42.1 to 15.3.
  • The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 77.8% for Grok-3 mini 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-3 miniLlama-3.3-70B-Instruct
ProviderxAIMeta
Noometry Index41.230.6
Released2025-04-092024-12-06
WeightsProprietaryOpen
Context window—128K
Max output—4K
Input $ / M tokens—$0.10
Output $ / M tokens—$0.32
Results tracked3543

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

Category by category

Coding Grok-3 mini leads

Grok-3 mini: 40.8 (#131), Llama-3.3-70B-Instruct: 31.0 (#290)

BenchmarkGrok-3 miniLlama-3.3-70B-Instruct
WeirdML42.6%14.4%
LMArena Coding13791268
Aider Polyglot49.3%—
SciCode—26%
BigCodeBench Instruct—46.9%
LiveBench Coding—36.6%
BigCodeBench Complete—57.5%

Agentic & Tool Use Not comparable

Grok-3 mini: —, Llama-3.3-70B-Instruct: 25.8 (#105)

BenchmarkGrok-3 miniLlama-3.3-70B-Instruct
Berkeley Function Calling Leaderboard—31.9%
BALROG—23%

Reasoning Too close to call

Grok-3 mini: 13.6 (#334), Llama-3.3-70B-Instruct: 14.1 (#327)

BenchmarkGrok-3 miniLlama-3.3-70B-Instruct
LMArena Hard Prompts13751257
Epoch Capabilities Index140.35127.33
ARC-AGI-20.4%—
SimpleBench—19.9%
Kagi LLM Benchmark61.3%—
ARC-AGI-116.5%—
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 mini leads

Grok-3 mini: 42.1 (#85), Llama-3.3-70B-Instruct: 15.3 (#298)

BenchmarkGrok-3 miniLlama-3.3-70B-Instruct
OTIS Mock AIME 2024-202577.8%5.1%
LMArena Math13861267
MATH Level 590.9%41.6%
Omni-MATH31.8%—
LiveBench Math—42.2%
FrontierMath (Feb 2025 set)5.9%—

Knowledge Grok-3 mini leads

Grok-3 mini: 46.4 (#81), Llama-3.3-70B-Instruct: 30.6 (#226)

BenchmarkGrok-3 miniLlama-3.3-70B-Instruct
GPQA Diamond76.3%47.4%
Confabulations10.8%22.8%
LMArena Expert13951225
MMLU-Pro79.9%—
Vectara Hallucination Rate—4.1%
GPQA (HELM)67.5%—
MMLU—86.3%

Multilingual Grok-3 mini leads

Grok-3 mini: 48.1 (#145), Llama-3.3-70B-Instruct: 39.9 (#220)

BenchmarkGrok-3 miniLlama-3.3-70B-Instruct
LMArena Non-English13521236
LMArena Chinese13871217
LMArena French13571281
LMArena German13491251
LMArena Japanese13421150
LMArena Korean13351143
LMArena Russian13531252
LMArena Spanish13811270

Instruction Following Grok-3 mini leads

Grok-3 mini: 78.5 (#9), Llama-3.3-70B-Instruct: 71.1 (#157)

BenchmarkGrok-3 miniLlama-3.3-70B-Instruct
LMArena Instruction Following13571242
LiveBench Instruction Following—82.7%
IFEval95.1%—

Long Context Grok-3 mini leads

Grok-3 mini: 41.0 (#147), Llama-3.3-70B-Instruct: 26.4 (#295)

BenchmarkGrok-3 miniLlama-3.3-70B-Instruct
Fiction.LiveBench66.7%33.3%
LMArena Longer Query13721256

Writing & Preference Grok-3 mini leads

Grok-3 mini: 52.5 (#169), Llama-3.3-70B-Instruct: 47.6 (#207)

BenchmarkGrok-3 miniLlama-3.3-70B-Instruct
LMArena Text13701274
LMArena Creative Writing13421250
LMArena Multi-Turn13551280
Short-Story Creative Writing73.5%—
WildBench65.1%—
LiveBench Language—39.2%

Frequently asked questions

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

Grok-3 mini is the stronger model overall, scoring 41.2 to 30.6 on the Noometry Index.

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

Grok-3 mini scores higher on coding benchmarks: 40.8 versus 31.0 in the Noometry coding category.

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

24 benchmarks have published results for both models. Grok-3 mini has 35 scored results on Noometry and Llama-3.3-70B-Instruct has 43.

Related comparisons

Go deeper