Unverified50% confidenceFactExact time
In the second phase, the Blender reference video, character reference sheet, and waterfront background image are all loaded together into the MiniMax H3 reference-image-to-video workflow in ComfyUI.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
9/26/2026
First Seen
Valid until: 12/25/2026
Sources
Related Entities
Related Claims
UnverifiedThe author uses the Minimax H3 default FL2VA first-and-last-frame video generation workflow in ComfyUI, setting the input image as both the first and last frame to form a closed-loop orbit shot.77% similarVerifiedAfter integration into ComfyUI, MiniMax H3 supports six core workflows: text-to-video, image-to-video, first-and-last-frame animation, reference-image-to-video, audio synchronization, and image editing.76% similarUnverifiedThis workflow is built in ComfyUI and requires three image-loading nodes (two character reference images and one scene image) all connected to the MiniMax H3 sampler.72% similarUnverifiedThe experiment uses a DiffusionGemma custom node paired with the LTX 2.3 video generation model, with inputs consisting of one static reference image, one reference video, and one natural language prompt.71% similarUnverifiedThe reference generation version of MiniMax H3 (REF VA) allows two or more reference images to be input to generate videos such as two-person dialogue videos.70% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/952757API
curl https://kongchang.com/api/v1/knowledge/claims/952757MCP
get_claim(id=952757)