Compare / Mistral Large 3 2512 vs GPT-5.6 Luna

Mistral Large 3 2512 vs GPT-5.6 Luna

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

Mistral Large 3 2512
Mistral · profile
GPT-5.6 Luna
OpenAI · profile
Results compared
19
Shared sources
3
BenchmarkMetricMistral Large 3 2512GPT-5.6 Luna
Arena (formerly LMArena)reportedOverallArena rating 1,4341,461
Arena (formerly LMArena)reportedBusiness, management and financeArena rating 1,4091,463
Arena (formerly LMArena)reportedCreative writingArena rating 1,3731,412
Arena (formerly LMArena)reportedExpert promptsArena rating 1,4281,491
Arena (formerly LMArena)reportedInstruction followingArena rating 1,4001,446
Arena (formerly LMArena)reportedOverallArena rating 1,4131,454
Arena (formerly LMArena)reportedWriting, literature and languageArena rating 1,3881,435
Artificial AnalysisreportedArtificial Analysis Coding Indexindex score 20.171.4
Artificial AnalysisreportedArtificial Analysis Intelligence Indexindex score 9.3037.3
Artificial AnalysisreportedGPQA Diamond% of questions 68.0%91.1%
Artificial AnalysisreportedHumanity's Last Exam% of questions 4.2%39.5%
Artificial AnalysisreportedLong-context reasoning (AA-LCR)% of questions 36.0%83.7%
Artificial AnalysisreportedSciCode% of problems 36.6%53.6%
Artificial AnalysisreportedTerminal-Bench 2.1% of tasks 12.0%80.9%
Artificial AnalysisreportedTerminal-Bench 4.0% of tasks 0.0%11.6%
Artificial AnalysisreportedΤ-bench banking% of tasks 5.8%31.1%
UGI LeaderboardreportedRequested-length error% off the requested word count 15.0%15.0%
UGI LeaderboardreportedStyle adherencescore from 0 to 1 0.380.38
UGI LeaderboardreportedWriting scorescore out of 100 41.653.7

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.