Noometry

Model comparison

GPT-5 vs Grok-3 mini

GPT-5 is the stronger model overall, scoring 50.9 to 41.2 on the Noometry Index.

Last verified . 35 shared benchmarks.

GPT-5 OpenAI

50.9

Rank #45 Confirmed

Grok-3 mini xAI

41.2

Rank #141 Confirmed

Summary

  • They share 35 benchmarks with published results for both. GPT-5 scores higher in 7 categories and Grok-3 mini in 1 category; 8 gaps are clear of the uncertainty.
  • The widest gap is in long context, where GPT-5 leads 69.5 to 41.0.
  • The biggest single-benchmark swing is ARC-AGI-1: 65.7% for GPT-5 and 16.5% for Grok-3 mini.

Side by side

GPT-5Grok-3 mini
ProviderOpenAIxAI
Noometry Index50.941.2
Released2025-08-072025-04-09
WeightsProprietaryProprietary
Context window400K—
Max output128K—
Input $ / M tokens$1.25—
Output $ / M tokens$10—
Results tracked6935

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

Category by category

Coding GPT-5 leads

GPT-5: 50.3 (#47), Grok-3 mini: 40.8 (#131)

BenchmarkGPT-5Grok-3 mini
Aider Polyglot88%49.3%
WeirdML60.7%42.6%
LMArena Coding14361379
SWE-bench Verified73.6%—
SWE-bench Verified (bash only)65%—
LMArena WebDev1418—
SciCode42.9%—
GSO6.9%—
ALE-Bench1,162—
AlgoTune1.67—

Agentic & Tool Use Not comparable

GPT-5: 33.1 (#56), Grok-3 mini: —

BenchmarkGPT-5Grok-3 mini
Terminal-Bench49.6%—
GDPval34.8%—
Remote Labor Index1.7%—
DeepResearch Bench49.6%—
BALROG32.8%—
LMArena Search1133—
METR Time Horizons69.6%—

Reasoning GPT-5 leads

GPT-5: 38.3 (#64), Grok-3 mini: 13.6 (#334)

BenchmarkGPT-5Grok-3 mini
ARC-AGI-29.9%0.4%
Kagi LLM Benchmark72.7%61.3%
ARC-AGI-165.7%16.5%
LMArena Hard Prompts14161375
Epoch Capabilities Index150140.35
SimpleBench56.7%—
CritPt12.6%—
Chess Puzzles37%—
EnigmaEval10.5%—
EBR-Bench12.7%—
Mystery Game Puzzles23%—
DTBench90.7%—
LMCA40%—
ForecastBench61.4—

Math GPT-5 leads

GPT-5: 55.0 (#44), Grok-3 mini: 42.1 (#85)

BenchmarkGPT-5Grok-3 mini
OTIS Mock AIME 2024-202591.4%77.8%
Omni-MATH64.7%31.8%
LMArena Math14071386
MATH Level 598.1%90.9%
FrontierMath (Feb 2025 set)32.4%5.9%
FrontierMath (Tiers 1-3)55.4%—
FrontierMath Tier 422%—
ProofBench18%—
FrontierMath Tier 4 (v1)12.5%—

Knowledge GPT-5 leads

GPT-5: 56.6 (#43), Grok-3 mini: 46.4 (#81)

BenchmarkGPT-5Grok-3 mini
GPQA Diamond86.2%76.3%
MMLU-Pro86.3%79.9%
Confabulations10.3%10.8%
GPQA (HELM)79.2%67.5%
LMArena Expert14191395
Humanity's Last Exam25.3%—
SimpleQA Verified50.1%—
Vectara Hallucination Rate14.7%—

Multimodal Not comparable

GPT-5: 46.8 (#13), Grok-3 mini: —

BenchmarkGPT-5Grok-3 mini
LMArena Vision1232—
GeoBench81%—
VPCT66%—

Multilingual GPT-5 leads

GPT-5: 51.4 (#110), Grok-3 mini: 48.1 (#145)

BenchmarkGPT-5Grok-3 mini
LMArena Non-English13971352
LMArena Chinese14221387
LMArena French14101357
LMArena German14161349
LMArena Japanese14091342
LMArena Korean13601335
LMArena Russian14061353
LMArena Spanish13991381

Instruction Following Grok-3 mini leads

GPT-5: 73.8 (#113), Grok-3 mini: 78.5 (#9)

BenchmarkGPT-5Grok-3 mini
IFEval87.5%95.1%
LMArena Instruction Following13881357

Long Context GPT-5 leads

GPT-5: 69.5 (#2), Grok-3 mini: 41.0 (#147)

BenchmarkGPT-5Grok-3 mini
Fiction.LiveBench97.2%66.7%
LMArena Longer Query13991372

Writing & Preference GPT-5 leads

GPT-5: 63.4 (#65), Grok-3 mini: 52.5 (#169)

BenchmarkGPT-5Grok-3 mini
LMArena Text14061370
LMArena Creative Writing13651342
Short-Story Creative Writing86%73.5%
WildBench85.7%65.1%
LMArena Multi-Turn14261355
EQ-Bench Creative Writing1627—

Frequently asked questions

Is GPT-5 better than Grok-3 mini?

GPT-5 is the stronger model overall, scoring 50.9 to 41.2 on the Noometry Index.

Is GPT-5 or Grok-3 mini better for coding?

GPT-5 scores higher on coding benchmarks: 50.3 versus 40.8 in the Noometry coding category.

How many benchmarks do GPT-5 and Grok-3 mini share?

35 benchmarks have published results for both models. GPT-5 has 69 scored results on Noometry and Grok-3 mini has 35.

Related comparisons

Go deeper