When using LiveAvatars in Custom Mode, as of now, we mainly handle the video generation that comes in from your audio. We do not handle the conversational elements between your app and your user.

This mode enables customers with pre-existing or custom LLM setups to directly plug their audio data into our systems and manage the output video on their own.

Configuration Options

As of now, we primarily offer the customization for:

• Avatars - control how the avatar looks
• LiveKit Configuration - control if you want to use our Livekit room or if you want us to send it into your room.

Avatars

An avatar determines the visual appearance of your AI agent. Developers can select from a catalog of avatars — ranging from realistic human presenters to stylized or expressive characters.

Each avatar can be uniquely identified by its avatar_id.

You can preview all available avatars here.

LiveKit

The LiveAvatar uses LiveKit to power our WebRTC set up. If you choose to, instead of using the room we provide by default, you can send us a LiveKit room information. We'll send our Avatar into your room instead.