Noometry

Model comparison

Grok-2 (Dec 2024) vs Grok 4

Grok 4 is the stronger model overall, scoring 48.1 to 33.7 on the Noometry Index.

Last verified . 25 shared benchmarks.

Grok-2 (Dec 2024) xAI

33.7

Rank #239 Confirmed

Grok 4 xAI

48.1

Rank #56 Confirmed

Summary

  • They share 25 benchmarks with published results for both. Grok-2 (Dec 2024) scores higher in 0 categories and Grok 4 in 8 categories; 8 gaps are clear of the uncertainty.
  • The widest gap is in math, where Grok 4 leads 48.4 to 20.8.
  • The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 11.5% for Grok-2 (Dec 2024) and 84% for Grok 4.

Side by side

Grok-2 (Dec 2024)Grok 4
ProviderxAIxAI
Noometry Index33.748.1
Released2024-08-132025-07-09
WeightsProprietaryProprietary
Context window——
Max output——
Input $ / M tokens——
Output $ / M tokens——
Results tracked3448

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

Category by category

Coding Grok 4 leads

Grok-2 (Dec 2024): 33.3 (#258), Grok 4: 50.3 (#46)

BenchmarkGrok-2 (Dec 2024)Grok 4
WeirdML22.2%45.7%
LMArena Coding12871408
Aider Polyglot—79.6%
LiveBench Coding46.4%—

Agentic & Tool Use Not comparable

Grok-2 (Dec 2024): —, Grok 4: 32.3 (#68)

BenchmarkGrok-2 (Dec 2024)Grok 4
Terminal-Bench—27.2%
Berkeley Function Calling Leaderboard—63%
GDPval—21.1%
Cybench—43%
DeepResearch Bench—47.3%
BALROG—43.6%
LMArena Search—1142
METR Time Horizons—66.6%

Reasoning Grok 4 leads

Grok-2 (Dec 2024): 16.9 (#299), Grok 4: 36.7 (#65)

BenchmarkGrok-2 (Dec 2024)Grok 4
SimpleBench22.7%60.5%
LMArena Hard Prompts12721409
Epoch Capabilities Index130.48146.44
ARC-AGI-2—16%
Kagi LLM Benchmark—73.6%
ARC-AGI-1—66.7%
Chess Puzzles—28%
LiveBench Reasoning54.8%—
DTBench65.2%—
LiveBench Data Analysis54.5%—
ForecastBench—60.9
LiveBench54.3%—

Math Grok 4 leads

Grok-2 (Dec 2024): 20.8 (#284), Grok 4: 48.4 (#64)

BenchmarkGrok-2 (Dec 2024)Grok 4
OTIS Mock AIME 2024-202511.5%84%
LMArena Math12831422
FrontierMath (Feb 2025 set)0.7%19.7%
Omni-MATH—60.3%
LiveBench Math54.9%—
MATH Level 563.5%—
FrontierMath Tier 4 (v1)—2.1%

Knowledge Grok 4 leads

Grok-2 (Dec 2024): 29.8 (#233), Grok 4: 53.8 (#55)

BenchmarkGrok-2 (Dec 2024)Grok 4
GPQA Diamond53.8%87%
Confabulations20.1%12.4%
LMArena Expert12541415
MMLU-Pro—85.1%
GPQA (HELM)—72.7%

Multimodal Not comparable

Grok-2 (Dec 2024): —, Grok 4: 33.7 (#94)

BenchmarkGrok-2 (Dec 2024)Grok 4
LMArena Vision—1210
GeoBench—45%

Multilingual Grok 4 leads

Grok-2 (Dec 2024): 43.1 (#188), Grok 4: 51.8 (#103)

BenchmarkGrok-2 (Dec 2024)Grok 4
LMArena Non-English12821403
LMArena Chinese12891427
LMArena French13181418
LMArena German12871429
LMArena Japanese12441394
LMArena Korean12371377
LMArena Russian12861410
LMArena Spanish12811420

Instruction Following Grok 4 leads

Grok-2 (Dec 2024): 66.9 (#202), Grok 4: 79.2 (#5)

BenchmarkGrok-2 (Dec 2024)Grok 4
LMArena Instruction Following12701387
LiveBench Instruction Following69.6%—
IFEval—94.9%

Long Context Grok 4 leads

Grok-2 (Dec 2024): 38.8 (#190), Grok 4: 63.1 (#4)

BenchmarkGrok-2 (Dec 2024)Grok 4
LMArena Longer Query12761409
Fiction.LiveBench—94.4%

Writing & Preference Grok 4 leads

Grok-2 (Dec 2024): 48.6 (#198), Grok 4: 58.5 (#116)

BenchmarkGrok-2 (Dec 2024)Grok 4
LMArena Text13051411
LMArena Creative Writing12841397
Short-Story Creative Writing63.6%76.9%
LMArena Multi-Turn12901416
WildBench—79.7%
LiveBench Language45.6%—

Frequently asked questions

Is Grok-2 (Dec 2024) better than Grok 4?

Grok 4 is the stronger model overall, scoring 48.1 to 33.7 on the Noometry Index.

Is Grok-2 (Dec 2024) or Grok 4 better for coding?

Grok 4 scores higher on coding benchmarks: 50.3 versus 33.3 in the Noometry coding category.

How many benchmarks do Grok-2 (Dec 2024) and Grok 4 share?

25 benchmarks have published results for both models. Grok-2 (Dec 2024) has 34 scored results on Noometry and Grok 4 has 48.

Related comparisons

Go deeper