Community ships 2 ComfyUI integrations for VDN-H3 within 48 hours of release
xiuyu_l · x · 2026-09-05
Just 48 hours after VDN-H3 was released for MiniMax's H3, the community shipped two independent ComfyUI integrations with 9 multimodal workflows. Although VDN was trained only on text-to-video, the community extended it to image-to-video and first-last-frame generation without retraining, validating how composable H3 is. The author notes this fulfills the video-gen dream that motivated their PhD work on diffusion efficiency.
More from Multimodal
- WIP lipsync test: ref2vid renders 15s of 2K video in 8 minutes — Kyrannio · 2026-09-05
- Bootstrapped Dev Says AI Video Test Runs Cost $1K Each, Consumers Won't Pay the Premium — Kyrannio · 2026-09-05
- Redditor shares daily ComfyUI H3 motion-context workflow with 7 node pack dependencies — Sad_Berry_4621 · 2026-09-05
- Astra stuns with engine-free game visuals, RocketLeagueBench declared saturated — SIGKITTEN · 2026-09-05
- GPT-6 generates a detailed Blender 3D model of Ilya Sutskever from a single photo — SIGKITTEN · 2026-09-05
- GPT-6 Astra Max beats GPT-5.6 Sol at visual essays and magazine spreads, user finds — dubesor · 2026-09-05