MiniMax H3 Video Model Goes Live with Versatile Reference Support

aziz4ai · x · 2026-08-03

Hailuo AI's MiniMax H3 video model is now live, featuring top-tier quality and versatile reference capabilities. A developer shared a workflow using the model: first using ChatGPT to create storyboards and motion graphics prompts, then uploading the image as a reference to MiniMax H3. By setting the dimensions to 16:9 and duration to 15 seconds, they successfully generated a high-quality Arabic motion-graphics film about smart money management.

Original post →

More from Multimodal

Multimodal channel →