From Personality to First Conversation
In the node-based editor, an Agent node holds the avatar's name and background, a Personality node picks one of 16 personality types, and a knowledge base holds the facts it answers from, such as a product sheet, a venue map or a course handbook. With a voice and the engine SDK on your model, the avatar makes the following possible:
- •The avatar listens through speech to text, with push to talk or voice detection, and speaks its reply as it is generated, in a designed or cloned voice.
- •The Face Animator in Unreal Engine and Godot lip syncs ARKit-style shapes, the Unity Audio Chat sample drives blendshape face animation, and every line carries one of 16 emotions for the brows, eyes and posture.
- •The action system and spatial awareness turn what people say into events, so the avatar can talk about and act on the objects you mark as interactable.
Group chat can put up to six characters in one conversation, with a director that picks who speaks next.
Keep the Avatar You Have
Neocortex is the mind, voice and lip sync, not an avatar maker, so you can bring a model from your own artists or an avatar creator. The character works in Unity, Unreal Engine, Godot and over the Web API, and the Web SDK adds text and voice chat to a website. Unity builds run on desktop, mobile and WebGL, and Unreal Engine on Windows, macOS, Android and iOS, for a guide in a virtual world, a face on a kiosk screen or a partner in VR.






























