Compare / Ministral 3 8B 2512 vs GPT-5.4

Ministral 3 8B 2512 vs GPT-5.4

5 results from 2 sources that measured both models. Each row is in the source's own units; there is no overall winner.

Ministral 3 8B 2512
Mistral · profile
GPT-5.4
OpenAI · profile
Results compared
5
Shared sources
2
BenchmarkMetricMinistral 3 8B 2512GPT-5.4
UGI LeaderboardreportedRequested-length error% off the requested word count 23.0%30.0%
UGI LeaderboardreportedStyle adherencescore from 0 to 1 0.330.40
UGI LeaderboardreportedWriting scorescore out of 100 28.868.8
Vectara Hallucination LeaderboardreportedAnswer rate% of documents 99.1%99.9%
Vectara Hallucination LeaderboardreportedHallucination rate% of summaries 21.7%7.0%

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.