Could anyone help me out to know which tts models ...
# agents
h
Could anyone help me out to know which tts models are supported from neuphonic in livekit ? I want to ask if this one is supported or not https://github.com/neuphonic/neutts-air
👀 1
@refined-appointment-81829 Could you please confirm this ?
l
This is a good question, also tagging @damp-kangaroo-50324 who may be able to provide some insight. I see
model
is documented at https://docs.livekit.io/agents/models/tts/plugins/neuphonic/ but the source does not accept a
model
parameter, https://github.com/livekit/agents/blob/main/livekit-plugins/livekit-plugins-neuphonic/livekit/plugins/neuphonic/tts.py#L62 (unless I'm seeing that wrong) It looks like the model chosen is dependent on the voice ID: https://docs.neuphonic.com/build-group/text-to-speech#text-to-speech-configuration
h
Hi @lively-winter-88668, Thanks for taking the time to look into this. Yes, that’s exactly my question - I was wondering why there isn’t a parameter to specify the model name. From what I’ve seen, it doesn’t seem like the model is automatically determined by the voice ID, so I suspect it might not be supported yet. That said, I could be wrong, and I’d really appreciate it if someone from your team could confirm this. My end goal is to use the new Neuphonic TTS Air model with LiveKit, so any guidance on how to make that work would be super helpful.
d
Hey guys! The livekit integration is for our prod model behind our API. We’re working on an integration for our OSS model but it’s not released yet.
Behind our API we have 2 models. We have our frontier model which is English only, and our smaller model for our non English languages. When our frontier model is multi lingual, we’ll deprecate our older multilingual model.