Overview

You can use custom classes in your functions to manage voice configuration and conversation states.

​

Conversation

The Conversation class is a core component of function execution, and is used for managing information during the conversation runtime.

def get_conversation_id(conv: Conversation):
    return conv.id

​

VoiceWeighting

Read the multi-voice document if you are unfamiliar with voice weighting.

The VoiceWeighting class allows you to assign specific weightings to voices, adjusting their prominence in a given context.

VoiceWeighting(
      voice=ElevenLabsVoice(
        provider_voice_id="LcfcDJNUP1GQjkzn1xUU",
        similarity_boost=0.2,
        stability=0.4
      ),
      weight=0.25
    )

​

ElevenLabsVoice

The ElevenLabsVoice class configures voice settings for ElevenLabs TTS system, allowing control over stability and similarity to the reference voice.

elevenlabs_voice = ElevenLabsVoice(
    provider_voice_id="a1b2C3d4E5f6G7h8I9j0",
    stability=0.5,
    similarity_boost=0.7
)

Parameters:

• provider_voice_id: Specifies the ElevenLabs voice to use. Specify using the ElevenLabs ID.
• stability: Controls the consistency of tone and delivery (0.0 to 1.0).
• similarity_boost: Controls how closely the generated voice matches the original (0.0 to 1.0).

​

PlayHTVoice

The PlayHTVoice class configures voice settings for PlayHT TTS system, allowing selection of a specific voice ID and style.

playht_voice = PlayHTVoice(
    provider_voice_id="en_us_male_1",
    style="conversational"
)

Parameters:

• provider_voice_id: Specifies the PlayHT voice to use.
• style: Defines the speaking style or tone.

​

RimeVoice

The RimeVoice class configures voice settings for Rime TTS system, allowing selection of a specific provider voice ID.

rime_voice = RimeVoice(
    provider_voice_id="s3://path-to-manifest.json",
    style="neutral",
    speed_alpha=1.0
)

Parameters:

• provider_voice_id: This mustb be an S3 path to a manifest file that contains the voice configuration.
• speed_alpha: Controls the speech rate multiplier. 1.0 is normal speed.

​

CartesiaVoice

The CartesiaVoice class configures voice settings for Cartesia TTS system.

cartesia_voice = CartesiaVoice(
    provider_voice_id="cartesia_ai_voice",
    pitch=1.1,
    rate=0.9,
    language="en",
    emotion="neutral"
)

Parameters:

• provider_voice_id: Specifies the Cartesia voice to use.
• speed: Controls the speed of speech.
• language: Sets the language code for synthesis (e.g., “en” for English).
• emotion: Allows setting the emotional tone of the speech.

​

CustomVoice

The CustomVoice class is used for defining voice configurations that use a custom TTS provider. This allows you to override the default system voice with a provider-supported voice.

voice_config = CustomVoice(provider="my_tts_provider", id="Emma")

Parameters:

• provider: Specifies the custom TTS provider.
• id: Specifies the voice ID within the custom provider.