Prompting Hailuo MiniMax: Testing the New Video Model with 12 Multimodal References

DavidmComfort · x · 2026-08-01

A user shared techniques for effectively prompting the latest Hailuo video generation model. The new model supports up to 12 multimodal references, allowing users to combine video, image, and audio inputs simultaneously.

Combined with text, this multimodal conditioning gives creators unprecedented control to bring complex storylines to life dynamically.

Original post →

More from Multimodal

Multimodal channel →