ElevenLabs KMP/dev.yveskalume.elevenlabs.tts/TextToSpeechApi

TextToSpeechApi

Generates speech from complete or incrementally produced text.

Functions

abstract suspend fun generate(request: TextToSpeechRequest): Audio

Generates a complete audio response for request.

abstract suspend fun openRealtimeSession(voiceId: String, options: RealtimeTtsOptions = RealtimeTtsOptions(), authorization: RealtimeTtsAuthorization = RealtimeTtsAuthorization.ConfiguredApiKey): RealtimeTtsSession

Opens a controllable realtime text-to-speech session for voiceId.

realtime

abstract fun realtime(voiceId: String, text: Flow<String>, options: RealtimeTtsOptions = RealtimeTtsOptions(), authorization: RealtimeTtsAuthorization = RealtimeTtsAuthorization.ConfiguredApiKey): Flow<AudioChunk>

Streams partial text input to ElevenLabs and emits audio as it becomes available. The returned flow is cold and opens a new realtime session for each collection. Empty input values are ignored, and normal input completion gracefully finishes the session.

stream

abstract fun stream(request: TextToSpeechRequest): Flow<AudioChunk>

Performs HTTP response streaming. The returned flow is cold and each collection starts a new request. Cancelling collection closes the underlying response. Each emitted AudioChunk is an arbitrary transport chunk, not necessarily a codec frame.