Text-to-speech entity

A text-to-speech (TTS) entity enables Home Assistant to speak to you.

A text-to-speech entity is derived from the homeassistant.components.tts.TextToSpeechEntity.

Properties

tip

Properties should always only return information from memory and not do I/O (like network requests).

Name	Type	Default	Description
supported_languages	list[str]	Required	The supported languages of the TTS service.
default_language	str	Required	The default language of the TTS service.
supported_options	list[str]	None	The supported options like voice, emotions of the TTS service.
default_options	Mapping[str, Any]	None	The default options of the TTS service.

Methods

Get supported voices

This method is used to return a list of supported voices for a language of a TTS service.

class MyTextToSpeechEntity(TextToSpeechEntity):
    """Represent a Text To Speech entity."""

    @callback
    def async_get_supported_voices(self, language: str) -> list[str] | None:
        """Return a list of supported voices for a language."""

Generating TTS audio in 1-shot

This method takes a message and language as input and returns the TTS audio. It can be implemented as either synchronous or asynchronous and is mandatory to implement.

class MyTextToSpeechEntity(TextToSpeechEntity):
    """Represent a Text To Speech entity."""

    def get_tts_audio(
        self, message: str, language: str, options: dict[str, Any]
    ) -> TtsAudioType:
        """Load tts audio file from the engine."""

    async def async_get_tts_audio(
        self, message: str, language: str, options: dict[str, Any]
    ) -> TtsAudioType:
        """Load tts audio file from the engine."""

Generating TTS audio with message streaming in

Large language models generate text in chunks. The TTS service can be called with a stream of text messages, and the TTS service will return the audio in chunks.

This method is optional. When not implemented, the TTS service will call the 1-shot method with the final message.

class MyTextToSpeechEntity(TextToSpeechEntity):
    """Represent a Text To Speech entity."""

    async def async_stream_tts_audio(
        self, request: TTSAudioRequest
    ) -> TTSAudioResponse:
        """Generate speech from an incoming message."""

The definition of the TTSAudioRequest and TTSAudioResponse objects are as follows:

@dataclass
class TTSAudioRequest:
    """Request to get TTS audio."""

    language: str
    options: dict[str, Any]
    message_gen: AsyncGenerator[str]


@dataclass
class TTSAudioResponse:
    """Response containing TTS audio stream."""

    extension: str
    data_gen: AsyncGenerator[bytes]