CustomAvatars
Override layer for premium 3D avatars, without touching core chat logic
Map a specific chat to a custom GLB model and the chat renders that instead of the default avatar. A thin, surgical extension point — invisible most of the time, irreplaceable for premium accounts.
- 1:1 Mapping from chat to custom model
- Auto Fallback to the standard library
- Thin Zero impact on core chat config
Most chats use the platform's standard 3D-avatar library and that is completely fine. Occasionally a business wants a unique character — a company mascot in 3D, or a licensed likeness. For those cases CustomAvatars exists.
The service is a mapping service, nothing more. It stores records that say "chat id X uses model file Y". When the chat opens on the customer's site, ChatBuilder asks CustomAvatars whether a custom model is mapped to this chat. If yes — that model is loaded. If no — the standard library is used.
Adding a custom model is not a self-service flow. New 3D assets require professional preparation, lip-sync rigging, and proper licensing. The operator team handles it — the model is uploaded to storage and a mapping record is created. The result is worth the care: a clean, well-rigged custom avatar looks and speaks beautifully.
The design choice is explicit. CustomAvatars is kept tiny so premium extensions don't pollute the core chat experience with extra settings the majority of users would never touch. The mapping layer is the clean seam between "standard" and "premium".
Model files are served in an optimised format so the customer's browser loads them quickly. The 3D viewer and lip-sync pipeline live in ChatBuilder; CustomAvatars only provides the asset. That separation is what makes the architecture extensible — new avatars can be added without changes in any other service.
Customers experience the result, not the mechanism. Lips sync to speech, gestures feel natural, the avatar embodies the brand. For businesses where that level of polish matters — luxury, concierge, entertainment — it is a differentiator worth the deliberate handling.
Everything CustomAvatars handles for you
-
Chat → model mapping
A per-chat override record that points to a custom 3D asset.
-
Automatic fallback
No mapping? Standard library is used transparently — zero config required.
-
Optimised asset delivery
GLB models served to browsers in an efficient format for fast load times.
-
Clean architectural seam
Premium-only feature isolated from core chat UI — no clutter for the 95%.
A thin extension behind ChatBuilder
CustomAvatars is called exclusively by ChatBuilder during the chat open sequence. If a mapping exists, it returns the asset reference; otherwise it says "not here" and the default flow proceeds.
- ChatBuilder queries CustomAvatars on every 3D-mode chat load and swaps in the premium asset if one is mapped.
Wire CustomAvatars into your product today
Book a consultation with our founders and we'll walk you through the whole microservice stack — not just this one — live on your domain.