SpeakshelfKOKORO TEXT TO SPEECH
Every Kokoro voice, on one page.
Play a sample of every voice of the open-weight Kokoro model. Filter by language and gender. No account, no setup, just the voices.
1 model family
Kokoro is a single model rather than a stable of engines, so its whole catalog sits in one family.
9 languages
Every language has its own page with the full list of voices, so you can link someone straight to one shelf of the catalog.
About this catalog
Kokoro is an open-weight text to speech model: 82 million parameters, released under the Apache 2.0 license. The AI TTS Microservice serves it alongside the cloud catalogs, so its 54 voices in 9 languages play here like any other shelf. Every voice has a playable sample.
Speakshelf is independent and not affiliated with Google, Amazon or the Kokoro project. Voice data and audio come from the AI TTS Microservice, a service that unifies Google, Amazon, Azure and other speech providers behind a single endpoint. Samples stream on demand, so listening is free.