Swap any video subject with one image: Viggle-Animate (MiniMax H3 finetune) needs no pose detection, 3 inference steps
cocktailpeanut · x · 2026-09-15
A developer demos Viggle-Animate, a MiniMax H3 finetune for video subject swapping: extract the first frame, use GPT Image 2.5 to swap characters, then feed that single reference image plus the original video into the model — it syncs the reference to the video till the end. No pose detection, no SAM, no segmentation pipeline; multiple subjects can be swapped at once, and it runs in just 3 inference steps. The author calls it criminally underrated.
Related event: Maestro Local AI Studio and Video Character Swap Workflows Gain Attention(3 posts)→
More from Multimodal
- HyperFrames + HeyGen: decent video from a basic prompt in ~10 minutes — toolstelegraph · 2026-10-30
- Hailuo MCP + Claude: chaining motion graphics generation through an agent — aziz4ai · 2026-10-30
- AI short film Nexus ep 7 cost $28,775 to make, creator shares full breakdown — NightsRadiant · 2026-10-30
- LayerShop: open-source Photoshop-style editor with ComfyUI local AI on layers — HonestTicket6013 · 2026-10-30
- AI video prompt goes viral: rally car racing a collapsing volcanic mountainside — umesh_ai · 2026-10-30
- Redditor finishes a short dinosaur movie made with Veo 3 and Seedance — HeadOpen4823 · 2026-10-30