//opengauntlet
← back to the leaderboard

Gemma 4 31B StyleTune (Q4_K_M) conversational AI benchmark

gemma4-31b-styletune-q4

Gemma 4 31B StyleTune (Q4_K_M) benchmark: humanlikeness scores, pairwise rankings, transcripts, latency, and measured local speed.

Leaderboard rank
#4 of 41
Humanlikeness
78.2 / 100
Pairwise rating
1771

Evidence: identical scenario pack and judge protocol; measured and judged results are shown separately below. Read the methodology.

Loading model…