Noometry

Model comparison

Grok 3 vs Grok 4.5

Grok 4.5 is the stronger model overall, scoring 55.0 to 39.9 on the Noometry Index.

Last verified . 26 shared benchmarks.

Grok 3 xAI

39.9

Rank #157 Confirmed

Grok 4.5 xAI

55.0

Rank #25 Confirmed

Summary

  • They share 26 benchmarks with published results for both. Grok 3 scores higher in 0 categories and Grok 4.5 in 9 categories; 8 gaps are clear of the uncertainty.
  • The widest gap is in reasoning, where Grok 4.5 leads 56.1 to 13.7.
  • The biggest single-benchmark swing is ARC-AGI-1: 5.5% for Grok 3 and 87.2% for Grok 4.5.

Side by side

Grok 3Grok 4.5
ProviderxAIxAI
Noometry Index39.955.0
Released2025-04-092026-07-08
WeightsProprietaryProprietary
Context window—500K
Max output—500K
Input $ / M tokens—$2
Output $ / M tokens—$6
Results tracked4052

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

Category by category

Coding Grok 4.5 leads

Grok 3: 41.9 (#115), Grok 4.5: 52.2 (#35)

BenchmarkGrok 3Grok 4.5
WeirdML37.2%46.4%
LMArena Coding14321474
DeepSWE—53.8%
FrontierCode—42.4%
Aider Polyglot53.3%—
LMArena WebDev—1553
SciCode—54.1%
ALE-Bench—1,309

Agentic & Tool Use Grok 4.5 leads

Grok 3: 30.5 (#76), Grok 4.5: 44.4 (#17)

BenchmarkGrok 3Grok 4.5
APEX-Agents—56.2%
τ²-bench Banking—47.9%
PostTrainBench—23.4%
BALROG29.5%—
GBAEval—65.4%
GDP.pdf—14%
LMArena Search—1213
Vending-Bench 2—3,887

Reasoning Grok 4.5 leads

Grok 3: 13.7 (#333), Grok 4.5: 56.1 (#25)

BenchmarkGrok 3Grok 4.5
ARC-AGI-20%52.6%
SimpleBench36.1%70%
Kagi LLM Benchmark61.3%83.5%
ARC-AGI-15.5%87.2%
LMArena Hard Prompts14341462
Epoch Capabilities Index138.33153.92
NYT Connections (extended)—79.9%
CritPt—15.4%
Chess Puzzles—36%
DTBench—96.5%
LMCA—45.2%
Surface Evolver Bench—74.4%

Math Grok 4.5 leads

Grok 3: 38.0 (#145), Grok 4.5: 60.9 (#35)

BenchmarkGrok 3Grok 4.5
OTIS Mock AIME 2024-202555.6%97.8%
LMArena Math13911459
FrontierMath (Tiers 1-3)—57.2%
FrontierMath Tier 4—24.4%
ProofBench—31%
Omni-MATH46.4%—
MATH Level 588.7%—
FrontierMath (Feb 2025 set)3.8%—
FrontierMath Tier 4 (v1)0%—

Knowledge Grok 4.5 leads

Grok 3: 46.2 (#82), Grok 4.5: 62.3 (#24)

BenchmarkGrok 3Grok 4.5
GPQA Diamond75.8%93.4%
LMArena Expert14211466
SimpleQA Verified—48.3%
MMLU-Pro78.8%—
Confabulations14.2%—
Vectara Hallucination Rate5.8%—
GPQA (HELM)65%—

Multimodal Not comparable

Grok 3: —, Grok 4.5: 37.6 (#72)

BenchmarkGrok 3Grok 4.5
LMArena Vision—1288
Blueprint-Bench 2—27.3%
Furniture Assembly—22.5%
LMArena Document—1452

Multilingual Grok 4.5 leads

Grok 3: 52.3 (#87), Grok 4.5: 54.4 (#42)

BenchmarkGrok 3Grok 4.5
LMArena Non-English14101440
LMArena Chinese14481496
LMArena French14601456
LMArena German14311446
LMArena Japanese13871428
LMArena Korean13731404
LMArena Russian14161448
LMArena Spanish14171450

Instruction Following Too close to call

Grok 3: 75.0 (#73), Grok 4.5: 76.0 (#48)

BenchmarkGrok 3Grok 4.5
LMArena Instruction Following14091446
IFEval88.4%—

Long Context Grok 4.5 leads

Grok 3: 38.7 (#192), Grok 4.5: 44.8 (#56)

BenchmarkGrok 3Grok 4.5
LMArena Longer Query14391463
Fiction.LiveBench58.3%—

Writing & Preference Grok 4.5 leads

Grok 3: 55.8 (#141), Grok 4.5: 65.8 (#42)

BenchmarkGrok 3Grok 4.5
LMArena Text14261448
LMArena Creative Writing14141442
EQ-Bench Creative Writing11861579
LMArena Multi-Turn14251456
Short-Story Creative Writing76.4%—
WildBench84.9%—

Frequently asked questions

Is Grok 3 better than Grok 4.5?

Grok 4.5 is the stronger model overall, scoring 55.0 to 39.9 on the Noometry Index.

Is Grok 3 or Grok 4.5 better for coding?

Grok 4.5 scores higher on coding benchmarks: 52.2 versus 41.9 in the Noometry coding category.

How many benchmarks do Grok 3 and Grok 4.5 share?

26 benchmarks have published results for both models. Grok 3 has 40 scored results on Noometry and Grok 4.5 has 52.

Related comparisons

Go deeper