Compare / Mistral Medium 3.5 vs GLM 5.3

Mistral Medium 3.5 vs GLM 5.3

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

Mistral Medium 3.5
Mistral · profile
GLM 5.3
Z.ai · profile
Results compared
25
Shared sources
3
BenchmarkMetricMistral Medium 3.5GLM 5.3
BulletBenchmeasured by usCost per gameUS dollars $0.0354$0.0156
BulletBenchmeasured by usGames lost on time% of games 0.0%83.3%
BulletBenchmeasured by usInvalid moves% of moves 0.4%0.0%
BulletBenchmeasured by usLadder Eloladder Elo 57233
BulletBenchmeasured by usMedian move timemilliseconds 0.5 s1.6 s
ROASBenchmeasured by usAudience scorescore out of 100 28.551.0
ROASBenchmeasured by usBusiness scorescore out of 100 3.412.9
ROASBenchmeasured by usConsistency scorescore out of 100 8.714.8
ROASBenchmeasured by usContribution profitsimulated US dollars −$159,882$9,718
ROASBenchmeasured by usCost of a runUS dollars $0.15$0.13
ROASBenchmeasured by usMonths over budgetmonths of 12 30
ROASBenchmeasured by usOverall scorescore out of 100 15.224.8
ROASBenchmeasured by usPlanning scorescore out of 100 55.251.7
ROASBenchmeasured by usReturn on ad spendprofit per $1 spent −$0.17$0.01
Arena (formerly LMArena)reportedConfirmed task successIPS effect estimate -0.160.06
Arena (formerly LMArena)reportedPraise over complaintIPS effect estimate -0.150.06
Arena (formerly LMArena)reportedSteerabilityIPS effect estimate -0.070.03
Arena (formerly LMArena)reportedTool groundingIPS effect estimate -0.020.00
Arena (formerly LMArena)reportedOverallArena rating 1,4301,468
Arena (formerly LMArena)reportedBusiness, management and financeArena rating 1,4301,474
Arena (formerly LMArena)reportedCreative writingArena rating 1,3981,456
Arena (formerly LMArena)reportedExpert promptsArena rating 1,4341,516
Arena (formerly LMArena)reportedInstruction followingArena rating 1,4211,480
Arena (formerly LMArena)reportedOverallArena rating 1,4261,480
Arena (formerly LMArena)reportedWriting, literature and languageArena rating 1,4041,470

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.