//opengauntlet
← back to the leaderboard

mythos-distilled-27b conversational AI benchmark

mythos-distilled-27b

mythos-distilled-27b benchmark: humanlikeness scores, pairwise rankings, transcripts, latency, and measured local speed.

Leaderboard rank
#32 of 41
Humanlikeness
56.7 / 100
Pairwise rating
1339
Measured speed
10.6 words/sec

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

Loading model…