Compare / Gemma 4 31B vs Grok 4.6

Gemma 4 31B vs Grok 4.6

21 results from 4 sources that measured both models. Each row is in the source's own units; there is no overall winner.

Gemma 4 31B
Google · profile
Grok 4.6
xAI · profile
Results compared
21
Shared sources
4
BenchmarkMetricGemma 4 31BGrok 4.6
Arena (formerly LMArena)reportedOverallArena rating 1,4451,461
Arena (formerly LMArena)reportedOverallArena rating 1,4501,449
Arena (formerly LMArena)reportedBusiness, management and financeArena rating 1,4491,447
Arena (formerly LMArena)reportedCreative writingArena rating 1,4191,445
Arena (formerly LMArena)reportedExpert promptsArena rating 1,4991,490
Arena (formerly LMArena)reportedInstruction followingArena rating 1,4531,451
Arena (formerly LMArena)reportedOverallArena rating 1,4531,453
Arena (formerly LMArena)reportedWriting, literature and languageArena rating 1,4351,446
Artificial AnalysisreportedArtificial Analysis Coding Indexindex score 43.476.8
Artificial AnalysisreportedArtificial Analysis Intelligence Indexindex score 14.744.3
Artificial AnalysisreportedGPQA Diamond% of questions 85.7%94.9%
Artificial AnalysisreportedHumanity's Last Exam% of questions 23.6%44.1%
Artificial AnalysisreportedLong-context reasoning (AA-LCR)% of questions 69.7%81.0%
Artificial AnalysisreportedSciCode% of problems 45.5%56.5%
Artificial AnalysisreportedTerminal-Bench 2.1% of tasks 43.4%88.4%
Artificial AnalysisreportedTerminal-Bench 4.0% of tasks 0.0%21.2%
Artificial AnalysisreportedΤ-bench banking% of tasks 14.8%50.7%
SimpleQA Verified (Epoch AI)reportedCorrect answers% of questions 10.4%49.3%
UGI LeaderboardreportedRequested-length error% off the requested word count 5.0%25.0%
UGI LeaderboardreportedStyle adherencescore from 0 to 1 0.360.36
UGI LeaderboardreportedWriting scorescore out of 100 43.163.3

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.