TTS 404s now name the fix instead of a generic not-found
POST /v1/audio/speech, /stream, and /stream/with-timestamps now return voice_not_found or a model-not-found message on a missing voice or model, instead of an opaque passthrough 404.
An unknown voice_id or model on POST /v1/audio/speech, POST /v1/audio/stream, or POST /v1/audio/stream/with-timestamps used to come back as a bare, opaque 404 - the same shape you’d get from any other not-found, with nothing telling you which one you’d hit.
All three endpoints now classify the cause: a bad voice_id returns the voice_not_found code with a message pointing you at GET /v1/voices; a bad model returns a not_found with a message pointing at GET /v1/audio/models. Nothing else about these endpoints’ error handling changed.
See the docs changelog entry for the full code/message table.