List available TTS models at runtime with GET /v1/audio/models
A new GET /v1/audio/models endpoint returns the models you can pass as the model parameter, each with a default and recommended flag plus the languages it supports — so a model picker can be built at runtime instead of hardcoding the list.
GET /v1/audio/models returns the text-to-speech models you can pass as the model parameter, so a model picker can be populated at runtime instead of hardcoding the list.
Each entry carries the model id, a human-readable name and description, a default flag (the model a request resolves to when it omits model), a recommended flag (what we suggest for new integrations, kept distinct from the default so backwards compatibility never shifts underneath you), and the languages it can synthesize as BCP-47 locale strings. These values reflect current support and change over time, so read them at runtime rather than caching them.
The Build changelog entry has the full response shape.