synthesize() - RunAnywhere Documentation

// Get the audio back
val audio = RunAnywhere.tts.synthesize("Hello from RunAnywhere")

// Play it
val handle = RunAnywhere.tts.speak("Hello from RunAnywhere")

Signatures

suspend fun synthesize(text: String, options: TtsOptions? = null): Audio
fun synthesizeStream(text: String, options: TtsOptions? = null): Flow<AudioChunk>
fun speak(text: String, options: TtsOptions? = null): SpeechHandle
suspend fun stop()
suspend fun voices(): List<Voice>

synthesize returns audio to save or feed elsewhere. speak plays it and returns a handle. speak is not suspending; it returns immediately.

Options

val audio = RunAnywhere.tts.synthesize(
    "Hello",
    TtsOptions(voice = "en_US-amy-medium", speed = 1.1f)
)

Field	Type	Default
`voice`	`String?`	none, model default
`language`	`String`	`DEFAULT_LANGUAGE`
`speed`	`Float`	`DEFAULT_SPEED`
`pitch`	`Float`	`DEFAULT_PITCH`
`format`	`AudioFormat`	`DEFAULT_FORMAT`
`sampleRate`	`Int`	`DEFAULT_SAMPLE_RATE`

All five defaults are public constants on TtsOptions, read from the proto defaults.

Stopping

RunAnywhere.tts.stop()

Stops whatever is speaking, without needing the handle.

Saving to a file

val audio = RunAnywhere.tts.synthesize("Save me")
File(context.filesDir, "out.wav").writeBytes(audio.samples)