Kokoro

Kokoro develops lightweight, efficient text-to-speech models optimized for natural-sounding voice synthesis.

Total Models

Text Models

Active Period

Dec 2024 – Dec 2024

Compare Kokoro Models

Kokoro 82M

Dec 2024

Lightweight 82M parameter TTS model based on StyleTTS2. The most popular text-to-speech model on Replicate with over 78 million runs. Known for natural-sounding speech with low latency.

text-to-speech

Model Evolution

See how Kokoro's models evolved by answering the same challenge across generations.