class TTS tts TTS def init self voice id str = DEFAULT VOICE LiveKit Community #ask-ai

class TTS(tts.TTS): def __init__( self...

narrow-hairdresser-82835

03/16/2026, 5:03 PM

class TTS(tts.TTS): def __init__( self, *, voice_id: str = DEFAULT_VOICE_ID, voice_settings: NotGivenOr[VoiceSettings] = NOT_GIVEN, model: TTSModels | str = "eleven_turbo_v2_5", encoding: NotGivenOr[TTSEncoding] = NOT_GIVEN, api_key: NotGivenOr[str] = NOT_GIVEN, base_url: NotGivenOr[str] = NOT_GIVEN, streaming_latency: NotGivenOr[int] = NOT_GIVEN, inactivity_timeout: int = WS_INACTIVITY_TIMEOUT, auto_mode: NotGivenOr[bool] = NOT_GIVEN, apply_text_normalization: Literal["auto", "off", "on"] = "auto", word_tokenizer: NotGivenOr[tokenize.WordTokenizer | tokenize.SentenceTokenizer] = NOT_GIVEN, enable_ssml_parsing: bool = Fa can we pass the voice settings like the above while using elevnlabs inference?

✅ 1

tall-airline-16756

03/16/2026, 5:04 PM

Thanks for reaching out, neyssi. I'm an AI assistant and I'm here to help. Give me a moment while I research this topic.

tall-airline-16756

03/16/2026, 5:04 PM

Thanks — you're asking whether you can pass detailed

voice_settings

(like

stability

similarity_boost

style

use_speaker_boost

) to the TTS constructor when using ElevenLabs via LiveKit Inference. Short answer: No — you cannot pass those ElevenLabs-specific

voice_settings

through LiveKit Inference. What you can do with LiveKit Inference: • Provide

model

(for example

elevenlabs/eleven_turbo_v2_5

) and

voice

(voice ID). • Optionally pass

language

. • Use

extra_kwargs

to forward a small set of additional ElevenLabs parameters such as

inactivity_timeout

apply_text_normalization

. What you cannot do via LiveKit Inference: • Customize ElevenLabs voice parameters like

stability

similarity_boost

style

, or

use_speaker_boost

. Those voice-level settings are only available when you use the ElevenLabs plugin directly with your own API key. If your goal is to tune those voice characteristics, the next step is to use the ElevenLabs plugin instead of LiveKit Inference. If you want, tell me which specific voice parameters you need (for example

stability

similarity_boost

) and I can show the exact plugin call/constructor you’d use or show how to map the options into

extra_kwargs

where applicable. Sources: ElevenLabs TTS | LiveKit Documentation, with interactive elements

8 Views

Open in Slack

Previous Next