JoyAI Echo Surgical Merge
This person seems to have improved/fixed JoyAI Echo (based on LTX 2.3) by improving character and voice consistency between scenes. Supposedly this is thus far the best solution for longform single-prompt multi-shot video generation.
Main Link: https://huggingface.co/joeygambino/joyai-echo-ltx23-echoVid-ltxAud-surgical
The person's entire huggingface has offshoot projects directly related to this.
I was just coming here to post this. I've tested his INT8 and it's surprisingly good at character and scene consistency with only a single sentence repeated. It's honestly better than anything else I've played with thus far.
He's still got a lot to work out in terms of nodes and workflows to make the experience smoother for the average user. Hoping RuneXX could help him on some of that.
Just started to play around with it as well. Its very resource heavy currently the nodes, hopefully some improvements there to come.
But will try figure out some way to use it all easily ;- )