Compare / Gemini 3.8 Flash vs Kimi K2.5

Gemini 3.8 Flash vs Kimi K2.5

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

Gemini 3.8 Flash
Google · profile
Kimi K2.5
Moonshot AI · profile
Results compared
19
Shared sources
5
BenchmarkMetricGemini 3.8 FlashKimi K2.5
Arena (formerly LMArena)reportedOverallArena rating 1,4771,445
Arena (formerly LMArena)reportedBusiness, management and financeArena rating 1,4821,441
Arena (formerly LMArena)reportedCreative writingArena rating 1,4821,423
Arena (formerly LMArena)reportedExpert promptsArena rating 1,5291,483
Arena (formerly LMArena)reportedInstruction followingArena rating 1,4881,439
Arena (formerly LMArena)reportedOverallArena rating 1,4921,450
Arena (formerly LMArena)reportedWriting, literature and languageArena rating 1,4861,428
Artificial AnalysisreportedArtificial Analysis Coding Indexindex score 76.346.8
Artificial AnalysisreportedArtificial Analysis Intelligence Indexindex score 40.923.5
Artificial AnalysisreportedGPQA Diamond% of questions 95.3%87.9%
Artificial AnalysisreportedHumanity's Last Exam% of questions 47.8%30.7%
Artificial AnalysisreportedLong-context reasoning (AA-LCR)% of questions 84.0%78.0%
Artificial AnalysisreportedTerminal-Bench 2.1% of tasks 87.6%45.7%
Artificial AnalysisreportedΤ-bench banking% of tasks 45.8%14.2%
SimpleQA Verified (Epoch AI)reportedCorrect answers% of questions 69.7%34.3%
UGI LeaderboardreportedRequested-length error% off the requested word count 13.0%10.0%
UGI LeaderboardreportedStyle adherencescore from 0 to 1 0.340.34
UGI LeaderboardreportedWriting scorescore out of 100 78.562.4
Vending-Bench 2reportedMoney after a yearUS dollars $5,093.79$1,198.46

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.