MiniMax H3 turns hand-drawn sketches into video effects

MiniMax's newly released video generation model H3 is drawing attention for a simple but striking capability: hand-drawn sketches can be turned directly into video effects. Across Chinese tech communities, the feature is being described as the multimodal field's Coding moment.

Based on public information, H3's core idea is to treat drawing as a natural input for video creation. Users no longer need to master complex post-production software — a few lines or shapes can be expanded by the model into full video scenes and effects, sharply lowering the barrier to video editing.

For creators, this shortens the distance from idea to finished footage. Work that once required editors and effects artists to collaborate for hours or days could now take a single sketch and one generation pass. The exclamation in coverage that video post-production is in danger reflects the industry's sense of disruption.

The Coding moment label matters because it evokes an earlier paradigm shift in programming: when simple inputs can drive complex outputs, professional skills become democratized and new groups of creators are unlocked. H3 aims to play the same role in video.

For MiniMax, H3 is a significant step in its video and multimodal product line. The company, whose consumer brand includes Hailuo AI, has been building a position in video generation, and H3 reinforces its push toward making creative tools accessible.

Of course, how the sketch-to-effect experience actually performs — in resolution, consistency, and fidelity — remains to be tested in real use. Whether the feature can enter professional post-production workflows will depend on the model's reliability in day-to-day work.

What to watch next: when H3 opens to the public, how it is priced, and whether rival video generation vendors follow with similar sketch-driven interfaces. Multimodal generation is moving from can it generate to can it generate well and controllably, and H3 is the latest sample on that path.