animede
Built from animede's public GitHub profile. Not on Take My Repo yet, so there are no posts or updates here.
MiniMax-H3 (video + synchronized audio, 33B) on diffusers — 9.7s per still, 44.2s per 5s clip on 48GB, with every speedup verified byte-identical
Speak into the mic, and the character answers back in real time — with voice and video.
Real-time talking character on a single RTX 5090 — LTX-2.5 22B + local LLM + TTS, fully local. First video ~2.6s after Enter.