Compare / DeepSeek V4 Pro 0423 vs Kimi K3

DeepSeek V4 Pro 0423 vs Kimi K3

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

DeepSeek V4 Pro 0423
DeepSeek · profile
Kimi K3
Moonshot AI · profile
Results compared
17
Shared sources
3
BenchmarkMetricDeepSeek V4 Pro 0423Kimi K3
BulletBenchmeasured by usLadder Eloladder Elo 00
BulletBenchmeasured by usMedian move timemilliseconds 6.4 s6.8 s
ROASBenchmeasured by usCost of a runUS dollars $0.20$1.01
ROASBenchmeasured by usMonths over budgetmonths of 12 00
ROASBenchmeasured by usOverall scorescore out of 100 28.244.1
ROASBenchmeasured by usReturn on ad spendprofit per £1 spent £0.26£0.50
Arena (formerly LMArena)reportedConfirmed task successIPS effect estimate -0.040.09
Arena (formerly LMArena)reportedPraise over complaintIPS effect estimate -0.020.07
Arena (formerly LMArena)reportedSteerabilityIPS effect estimate -0.03-0.01
Arena (formerly LMArena)reportedTool groundingIPS effect estimate -0.000.00
Arena (formerly LMArena)reportedOverallArena rating 1,4551,472
Arena (formerly LMArena)reportedBusiness, management and financeArena rating 1,4601,489
Arena (formerly LMArena)reportedCreative writingArena rating 1,4461,458
Arena (formerly LMArena)reportedExpert promptsArena rating 1,4801,533
Arena (formerly LMArena)reportedInstruction followingArena rating 1,4521,486
Arena (formerly LMArena)reportedOverallArena rating 1,4581,488
Arena (formerly LMArena)reportedWriting, literature and languageArena rating 1,4511,475

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.