미니맥스 H3

What is 미니맥스 H3?

MiniMax H3, also known as Hailuo 3.0, is an omnimodal video generation model that understands text, images, video, and audio together. It lets you combine up to 9 reference images, 3 videos, and 3 audio files to create 4 to 15 second videos with native stereo sound and up to 2K resolution. The product is built for creative control, using an @ reference system to assign different roles to each asset, like character style, camera motion, or soundtrack. It supports a wide range of use cases such as ads, brand campaigns, ecommerce, product design, UI/UX, gaming, video editing, and cinematic storytelling. It also emphasizes consistency across shots, motion transfer, long takes, video extension, and character replacement. A free mode is available for text-to-video or image-to-video, while subscribers get full multimodal creation. The site says the model can be used through hailuoai.video and the MiniMax Open Platform API.

Key features

• Combines text, images, video, and audio in one generation flow.
• Supports up to nine images, three videos, and three audio files.
• Creates 4 to 15 second videos with native stereo sound.
• Uses @ references to control each asset’s role precisely.
• Maintains character, product, and text consistency across shots.
• Handles video editing, motion transfer, extension, and long takes.