Ten free talking-face engines, one photo, one drawing and one cruel sentence about a man named Bob. What each one did on our RTX 5090, what the licences say, why our lip metric picked the wrong winners, how SoulX-FlashHead Pro got fast enough to go live, and the memory leak we caught before it could eat the PC.
For five weeks we would not say how our presenter’s talking footage is made. Here is the whole chain: LongCat for the body, MuseTalk for the mouth, Duix to clean up what that leaves, and MuseTalk again for the visemes. Why the order matters more than any single model.
Duix (formerly HeyGem) is the open-source HeyGen clone, it clones a face from a short video and re-lip-syncs it to any audio, offline on an NVIDIA card. We installed it and ran it. Here’s the honest result.
InfiniteTalk promises infinite-length talking video from a portrait and audio, running in ComfyUI. It looks very good, but on a 32GB card the ‘infinite’ part isn’t practical yet. Here’s the honest state.
LongCat animates a whole body, gestures, lean, face, mouth, from audio, locally, for free. Long takes in one job. It’s our pick for a full-body local host, with two honest caveats: a quiet background, and drift on very long takes.
LTX’s native audio dub is the highest-resolution lip-sync we’ve run locally, but the generative mode invents a made-up language unless you wire the true audio-driven path. Here’s how it works and where it slips.
Sonic makes lips sync from an image and audio, but it regenerates the whole face every frame, so the eyes roll, irises drift, and teeth get reinvented. Here’s why it’s unshippable for a recurring character.