jambonz / jambonz-webapp

A simple provisioning web app for jambonz
MIT License
5 stars 20 forks source link

Add ElevenLabs as TTS vendor #328

Closed davehorton closed 9 months ago

davehorton commented 9 months ago

https://elevenlabs.io/speech-synthesis

davehorton commented 9 months ago

Eleven Labs list of voices here.

ElevenLabs does not have a separate list of voices per language. Instead, each language they support has the same list of voices. Getting the list of languages they support is slightly different than other synthesizers. They support a few "models" and each model has a list of languages that is supported. THe current models are:

To start with we should offer all of the languages supported by the eleven_multilingual_v2 model. Details on the models can be found here, including the list of languages.

So for the Applications page, when someone selects ElevenLabs for speech synthesis, show all 29 languages that are supported by the eleven_multilingual_v2 model. The language dropdown should show the language name (e.g. English) and store the associated code as defined in the model as language_id (e.g. en).

Regardless of the language selected, the same list of voices should be shown. Per the list of voices in the link above, each voice has a name (Rachel) and a voice_id (21m00Tcm4TlvDq8ikWAM). The voice id is what should be stored in the database.

davehorton commented 9 months ago

For testing create a free account https://elevenlabs.io/sign-up