lobe-vidol

Why we included this project

Instead of returning AI replies as a plain text box, this project renders dialogue through a live 3D avatar in the browser, so the conversation feels like it comes from an animated character you can pose and set dancing with MMD motion files. It is a solid working example for anyone building AI companions, virtual idols, or VTuber-style experiences: the Next.js frontend handles text chat, voice, and 3D animation together and connects to multiple model providers. The character editor and shared marketplace let you upload and reuse community-made VRM models instead of starting from scratch. Keep in mind it is early-stage and still in beta, so treat it as a strong prototype reference rather than a finished production system.