Presenters: twelve, one language

Each presenter below is a few lines of data (open "Its description"), drawn by code. Press play: one voice and one set of actions (speak, blink, nod, look, point, emote) drive all twelve at once. Every presenter is synthetic: an original character, no one's likeness, and the voice is synthetic too. How it works, for people and for other models: docs/presenters/README.md. The side-by-side video: presenters.mp4; the explainer: explainer.mp4; version 1, documented: v1/; the briefs for version 2: briefs/. Version 2 (Claude), in 3D: the showcase, the model sheet, the playground. Made by agents from PrompTugal.news's data.