MiniMax H3

About MiniMax H3

MiniMax H3 is a web app for generating short videos using the MiniMax H3 video model. Users can create videos in two ways: text-to-video by describing a scene in a prompt, or image-to-video by uploading a first and/or last frame to guide the generation. Videos can be 4 to 15 seconds long at 768P or 2K resolution.

The service supports text-to-video aspect ratios of 21:9, 16:9, 4:3, 1:1, 3:4, and 9:16. Users can upload JPEG, PNG, or WebP images up to 5 MB with dimensions of 256–5760 pixels per side. Prompts support up to 7,000 characters and can describe the subject, setting, motion, camera movements, lighting, mood, and audio elements.

Generation credits are calculated by resolution and duration, reserved when a task is submitted, and automatically released if generation fails. Content screening prevents sexual content, material involving minors, real person likenesses without consent, and content infringing others' rights. The service is built with documented limits, clear settings, and responsible use policies.

• web-app
• prompt-based
• image to video
• video-generation
• 2k video
• 768p
• ai video
• text to video

FAQ

What is h3minimax.com?

A web app for creating short videos with the MiniMax H3 video model, from a text prompt or a first and last frame.

What can I create with this tool?

Short videos from a text prompt (text-to-video), or from a first and/or last frame image plus a prompt (image-to-video). Image generation, video-to-video, and video editing are not offered.

How are credits charged?

Credits are calculated on the server from the resolution and duration, reserved when a task is accepted, and released automatically if the generation fails.

What content is not allowed?

Sexual content, any content involving minors, a real person's likeness without their consent, and content that infringes others' rights are prohibited. Prompts and uploads are screened before generation, and violations lead to account suspension.

Which lengths and resolutions are supported?

Any whole-number duration from 4 to 15 seconds, at 768P or 2K. Text-to-video supports 21:9, 16:9, 4:3, 1:1, 3:4, and 9:16 aspect ratios; image-to-video follows the uploaded frame.