Discussions

Hello everyone,

I am currently integrating LiveAvatar into a production environment using FULL Mode via the Python SDK and LiveKit. While the integration is working well, I’ve encountered a few challenges regarding the avatar's visual behavior and would appreciate some guidance:

Idle Behavior (Blinking): When the avatar is listening or in an idle state, it appears completely static—it doesn't even blink. This makes the interaction feel a bit unnatural. Is there a specific parameter in the avatar_persona configuration or a LiveKit event I can send to enable "idle" animations like blinking or subtle breathing?

Animation Depth (Head vs. Mouth movements): I’ve noticed that during speech, sometimes only the mouth moves while the head remains perfectly still. At other times, there is full-head movement. What triggers this difference? Can I influence the "intensity" of the movement (e.g., a "natural" vs. "static" mode) via the API payload or the system prompt?

Influence via Code: Besides the standard avatar_id and voice_id, are there any undocumented or advanced parameters in the avatar_config that allow for more control over the avatar's non-verbal expressions?

I am currently sending text Satz-für-Satz (sentence-by-sentence) and would love to know if there's a way to keep the avatar "alive" during the gaps between responses.

Thank you in advance for your help!

Best regards