MiniMax H3 / Hailuo 3.0 · Ref2VA

One Character Moves through Multiple Cities

The same woman appears across changing cities, outfits, and lighting, emphasizing reference-character consistency. Native-video output: 15s · 928×1664. The source did not publish a complete prompt, so this library does not infer or complete one.

Generation mode
Ref2VA
Model
MiniMax H3 (creator-labeled)
Output
15s · 928×1664 · portrait
Tags
Local Generation & Workflow · Photoreal · Vertical short-form · Fashion · City & street
Prompt provenance
Not published by the source

Prompt not published by the source
This page only shows the video and verifiable public information. It does not infer or complete a prompt from the video.

View original source · Original X post · @ai_nontan_room