Compare / Gemma 3 12B vs grok-4-0709

Gemma 3 12B vs grok-4-0709

17 results from 3 sources that measured both models. Each row is in the source's own units; there is no overall winner.

Gemma 3 12B
Google · profile
grok-4-0709
xAI · profile
Results compared
17
Shared sources
3
BenchmarkMetricGemma 3 12Bgrok-4-0709
Arena (formerly LMArena)reportedBusiness, management and financeArena rating 1,3861,395
Arena (formerly LMArena)reportedCreative writingArena rating 1,3321,395
Arena (formerly LMArena)reportedExpert promptsArena rating 1,2741,431
Arena (formerly LMArena)reportedInstruction followingArena rating 1,3211,397
Arena (formerly LMArena)reportedOverallArena rating 1,3421,411
Arena (formerly LMArena)reportedWriting, literature and languageArena rating 1,3131,399
Berkeley Function Calling Leaderboard (BFCL) V4reportedIrrelevance detection% correct 70.3%84.3%
Berkeley Function Calling Leaderboard (BFCL) V4reportedMemory% correct 27.5%55.9%
Berkeley Function Calling Leaderboard (BFCL) V4reportedMulti-turn tasks% correct 5.8%47.0%
Berkeley Function Calling Leaderboard (BFCL) V4reportedOverall accuracy% correct 30.4%63.0%
Berkeley Function Calling Leaderboard (BFCL) V4reportedRelevance detection% correct 93.8%87.5%
Berkeley Function Calling Leaderboard (BFCL) V4reportedSingle-turn calls (curated)% correct 79.4%85.4%
Berkeley Function Calling Leaderboard (BFCL) V4reportedSingle-turn calls (user-contributed)% correct 74.2%75.6%
Berkeley Function Calling Leaderboard (BFCL) V4reportedWeb search% correct 4.0%82.0%
UGI LeaderboardreportedRequested-length error% off the requested word count 9.0%17.0%
UGI LeaderboardreportedStyle adherencescore from 0 to 1 0.290.36
UGI LeaderboardreportedWriting scorescore out of 100 29.967.0

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.