🎬 JoyAI-Echo × LTX-2.3 — Multi-Shot Narrated Video

A surgical merge of JoyAI-Echo (cross-shot character memory) and LTX-2.3-distilled (natural voice + lip-sync) for joint audio-video generation. Write a story as one or more shots (separate shots with a line containing only ---); the paired memory bank keeps the same character's face and voice consistent across shots.

Model: joeygambino/joyai-echo-ltx23-echoVid-ltxAud-surgical · fp8 checkpoint · text encoder: Gemma-3-12B · non-commercial (LTX-2 Community License). Generated content is machine-generated.

Examples