Compare / Claude Sonnet 4.5 vs Grok 4.6

Claude Sonnet 4.5 vs Grok 4.6

22 results from 5 sources that measured both models. Each row is in the source's own units; there is no overall winner.

Claude Sonnet 4.5
Anthropic · profile
Grok 4.6
xAI · profile
Results compared
22
Shared sources
5
BenchmarkMetricClaude Sonnet 4.5Grok 4.6
Arena (formerly LMArena)reportedOverallArena rating 1,4491,461
Arena (formerly LMArena)reportedOverallArena rating 1,4491,449
Arena (formerly LMArena)reportedBusiness, management and financeArena rating 1,4611,447
Arena (formerly LMArena)reportedCreative writingArena rating 1,4541,445
Arena (formerly LMArena)reportedExpert promptsArena rating 1,4981,490
Arena (formerly LMArena)reportedInstruction followingArena rating 1,4641,451
Arena (formerly LMArena)reportedOverallArena rating 1,4571,453
Arena (formerly LMArena)reportedWriting, literature and languageArena rating 1,4541,446
Artificial AnalysisreportedArtificial Analysis Coding Indexindex score 52.176.8
Artificial AnalysisreportedArtificial Analysis Intelligence Indexindex score 20.744.3
Artificial AnalysisreportedGPQA Diamond% of questions 83.4%94.9%
Artificial AnalysisreportedHumanity's Last Exam% of questions 17.8%44.1%
Artificial AnalysisreportedLong-context reasoning (AA-LCR)% of questions 72.3%81.0%
Artificial AnalysisreportedSciCode% of problems 45.7%56.5%
Artificial AnalysisreportedTerminal-Bench 2.1% of tasks 55.8%88.4%
Artificial AnalysisreportedTerminal-Bench 4.0% of tasks 0.0%21.2%
Artificial AnalysisreportedΤ-bench banking% of tasks 24.5%50.7%
SimpleQA Verified (Epoch AI)reportedCorrect answers% of questions 30.7%49.3%
UGI LeaderboardreportedRequested-length error% off the requested word count 5.0%25.0%
UGI LeaderboardreportedStyle adherencescore from 0 to 1 0.410.36
UGI LeaderboardreportedWriting scorescore out of 100 66.963.3
Vending-Bench 2reportedMoney after a yearUS dollars $3,838.74$9,047.03

Bold green marks the better value on that metric. Where a source reports ranges that overlap, the difference may not be meaningful; see the benchmark page for ranges.