Shared concepts
Language mixing
How Soniox Text-to-Speech handles multiple languages within a request or session.
Overview
Soniox Text-to-Speech supports 60+ languages, but each request or
session uses a single language, set through the language field.
Mixing languages in one request
Mixing multiple languages within the same request or session is not supported yet. This capability will be released in the coming weeks.
Current solution
To generate speech in multiple languages, split the text by language and send a separate request or session for each one. You can then combine the resulting audio on your side.
For example, to produce an English sentence followed by a French one:
- Send the English text with
"language": "en". - Send the French text with
"language": "fr". - Concatenate the two audio outputs.
Using the same voice across requests keeps the speaker consistent. All Soniox voices work with every supported language. See voices for the full list.