Noometry

Model comparison

Claude 3 Opus vs Grok 4.3

Grok 4.3 is the stronger model overall, scoring 43.8 to 29.5 on the Noometry Index.

Last verified . 27 shared benchmarks.

Claude 3 Opus Anthropic

29.5

Rank #310 Confirmed

Grok 4.3 xAI

43.8

Rank #86 Confirmed

Summary

  • They share 27 benchmarks with published results for both. Claude 3 Opus scores higher in 0 categories and Grok 4.3 in 10 categories; 10 gaps are clear of the uncertainty.
  • The widest gap is in math, where Grok 4.3 leads 46.0 to 14.8.
  • The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 4.7% for Claude 3 Opus and 93.3% for Grok 4.3.

Side by side

Claude 3 OpusGrok 4.3
ProviderAnthropicxAI
Noometry Index29.543.8
Released2024-02-292026-04-17
WeightsProprietaryProprietary
Context window—1M
Max output—30K
Input $ / M tokens—$1.25
Output $ / M tokens—$2.50
Results tracked4640

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

Category by category

Coding Grok 4.3 leads

Claude 3 Opus: 32.9 (#267), Grok 4.3: 41.6 (#121)

BenchmarkClaude 3 OpusGrok 4.3
WeirdML19.2%49.9%
LMArena Coding12641415
LMArena WebDev—1357
SciCode—47.3%
BigCodeBench Instruct45.5%—
LiveBench Coding38.6%—
BigCodeBench Complete57.4%—
ALE-Bench—944.17
HumanEval+77.4%—
MBPP+73.3%—

Agentic & Tool Use Grok 4.3 leads

Claude 3 Opus: 24.6 (#116), Grok 4.3: 27.7 (#99)

BenchmarkClaude 3 OpusGrok 4.3
Cybench10%—
GDP.pdf—8%
LMArena Search—1165
METR Time Horizons29.5%—
Vending-Bench 2—35.26

Reasoning Grok 4.3 leads

Claude 3 Opus: 14.6 (#324), Grok 4.3: 35.9 (#68)

BenchmarkClaude 3 OpusGrok 4.3
Chess Puzzles5%25%
LMArena Hard Prompts12451396
DTBench61.6%90.7%
LMCA17%38.3%
Epoch Capabilities Index126.91149.16
ForecastBench58.460.3
SimpleBench23.5%—
NYT Connections (extended)—55.2%
CritPt—8%
EnigmaEval0.8%—
LiveBench Reasoning40.6%—
LiveBench Data Analysis57.9%—
LiveBench49.2%—
WinoGrande88.5%—

Math Grok 4.3 leads

Claude 3 Opus: 14.8 (#299), Grok 4.3: 46.0 (#74)

BenchmarkClaude 3 OpusGrok 4.3
OTIS Mock AIME 2024-20254.7%93.3%
LMArena Math12731388
FrontierMath (Tiers 1-3)—42.8%
FrontierMath Tier 4—14.6%
ProofBench—11%
LiveBench Math43.6%—
MATH Level 537.5%—

Knowledge Grok 4.3 leads

Claude 3 Opus: 24.5 (#267), Grok 4.3: 52.5 (#62)

BenchmarkClaude 3 OpusGrok 4.3
GPQA Diamond47.2%88.8%
SimpleQA Verified12.6%33.2%
LMArena Expert12231385
Confabulations22.7%—
MMLU84.6%—

Multimodal Grok 4.3 leads

Claude 3 Opus: 27.1 (#116), Grok 4.3: 31.6 (#104)

BenchmarkClaude 3 OpusGrok 4.3
LMArena Vision10231229
Blueprint-Bench 2—0%

Multilingual Grok 4.3 leads

Claude 3 Opus: 41.4 (#207), Grok 4.3: 50.5 (#120)

BenchmarkClaude 3 OpusGrok 4.3
LMArena Non-English12581385
LMArena Chinese12481422
LMArena French12751412
LMArena German12581395
LMArena Japanese12041379
LMArena Korean11871356
LMArena Russian12801399
LMArena Spanish12461398

Instruction Following Grok 4.3 leads

Claude 3 Opus: 64.1 (#228), Grok 4.3: 72.1 (#140)

BenchmarkClaude 3 OpusGrok 4.3
LMArena Instruction Following12481366
LiveBench Instruction Following63.9%—

Long Context Grok 4.3 leads

Claude 3 Opus: 38.2 (#202), Grok 4.3: 42.5 (#123)

BenchmarkClaude 3 OpusGrok 4.3
LMArena Longer Query12591393

Writing & Preference Grok 4.3 leads

Claude 3 Opus: 47.2 (#213), Grok 4.3: 58.5 (#118)

BenchmarkClaude 3 OpusGrok 4.3
LMArena Text12621397
LMArena Creative Writing12351380
LMArena Multi-Turn12751406
EQ-Bench 4—1075
LiveBench Language50.4%—

Frequently asked questions

Is Claude 3 Opus better than Grok 4.3?

Grok 4.3 is the stronger model overall, scoring 43.8 to 29.5 on the Noometry Index.

Is Claude 3 Opus or Grok 4.3 better for coding?

Grok 4.3 scores higher on coding benchmarks: 41.6 versus 32.9 in the Noometry coding category.

How many benchmarks do Claude 3 Opus and Grok 4.3 share?

27 benchmarks have published results for both models. Claude 3 Opus has 46 scored results on Noometry and Grok 4.3 has 40.

Related comparisons

Go deeper