The open-source text-to-speech model Kokoro 82M v1.0 has been released, positioning itself as a leading downloadable option. While it performs well, it trails behind ElevenLabs' Eleven v4 Turbo by 264 ELO points in blind human preference tests. AI
IMPACT Provides a new open-source option for text-to-speech, though it currently lags behind commercial offerings in human preference.
RANK_REASON Release of an open-source model with benchmark comparisons. [lever_c_demoted from research: ic=1 ai=1.0]
Read on Mastodon — mastodon.social →
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →