ComfyUI-Ref2VA-VSA
Type: ComfyUICategory: Generation★ 524 (GitHub, when we checked)Spotted, not yet tested
ComfyUI-Ref2VA-VSA is a ComfyUI workflow or node pack in our Generation category. Sparse-attention MiniMax H3 reference-to-video node: README reports ~72s warm for a 5s 1344×768 clip on an RTX 4090, 4 steps, ~13.5 GB peak VRAM, with reference tokens kept out of pruning. Apache-2.0.
Where it fits: ComfyUI, Open-weight video & audio models.
Listing isn’t endorsement. This is our summary of the project’s own docs, not a Studio Hall test result. Check the licence and requirements at the source before you build on it.
Lab posts that cover ComfyUI-Ref2VA-VSA
- MiniMax H3 gets fast locally: Sol-H3 Studio, HyperFlow 8-step, Ref2VA-VSASpotted, not yet tested
Related Lab posts
- FastH3 Omni goes open: reference-to-video with sound on your own GPU, and the case for letting Opus cut cheap clipsSpotted, not yet tested
- ComfyArt: a local Lovart-style infinite canvas on your ComfyUISpotted, not yet tested
- Official ComfyUI: Blender blockout to Seedance 2.5Spotted, not yet tested
- #OpenSourceHiggsfield: Comfy's $25K challenge and the first open rebuildsSpotted, not yet tested
- Fish Audio Drama 3 in ComfyUI: directed multi-speaker voice, then H3 talking videoBased on the creator’s video · not yet tested by us
Similar skills & repos
- FastH3 Omni · ComfyUI, Generation
- seriesmith · ComfyUI, Generation
- Comfield · Repo, Generation
- Sol-H3 in ComfyUI · ComfyUI, Generation
- ComfyUI Blender→Seedance 2.5 · ComfyUI, Generation
- ComfyUI Wan 3.0 templates · ComfyUI, Generation