H3Shot DeskRun planner Search/
All H3 reference notes
Mode selection

When should you use Reference Generation?

Answer

Use Reference Generation when several images, videos, or audio clips need distinct roles such as identity, wardrobe, setting, motion, or sound. It is unnecessary for a scene text alone can define.

Try this sequence
  1. Write one role beside every input.
  2. Remove duplicate references with conflicting details.
  3. State those roles before the shot description.
Plan this workflowOpen MiniMax sourceVerified against MiniMax docs / Aug 5, 2026
Related problemsQText-to-Video or Image-to-Video?Use Text-to-Video when the scene can be invented. Use First/Last-Frame Image-to-Video when the opening composition, subject, product, or interface must begin or end from a supplied image.Mode selectionQFirst frame or first-and-last frame?Use First/Last-Frame Image-to-Video with one image to animate away from a controlled start. Use two images when both endpoint compositions matter and H3 must create the transition.Mode selection