MiniMax H3 Lite Text-to-Video Model

Overview

MiniMax H3 Lite is a video generation model made by MiniMax. It turns a written scene description into a short MP4 video clip. You can also anchor the motion by supplying a still first frame, and optionally a still last frame. This helps you keep composition stable while the model animates the shot.

What you can build

• Storyboard clips and animatics from text descriptions
• Product shots with controlled openings and endings using still frames
• Social-ready vertical clips by choosing a 9:16 aspect ratio
• Simple camera moves like push-ins, pans, and tracking shots described in text
• Character or subject reveals that start on a specific frame and end on a target frame

Inputs

• A required written prompt that describes the scene, motion, camera, and style. If you include dialogue in quotes, the model may attempt to speak it on camera.
• An optional first-frame image upload that sets the opening frame of the video. WebP images are supported. When you provide a first frame, the output follows that image’s aspect ratio.
• A required clip length in whole seconds. It must be between 4 and 15 seconds.
• A required output resolution choice. Pick 480P for smaller files or 768P for sharper results.
• An optional last-frame image upload that sets the ending frame. This is most useful when you also provide a first frame.
• An optional aspect ratio for text-to-video clips. Supported ratios include 21:9, 16:9, 4:3, 1:1, 3:4, and 9:16. If you provide a first frame image, the image shape takes priority.

Outputs

The model returns a single generated video as an MP4 file. The clip matches the requested duration and uses the selected resolution. When you drive the generation with a first frame image, the output keeps that image’s aspect ratio. Depending on your prompt, the video can include synchronized audio alongside the visuals.

Limitations

• Output duration is limited to 4–15 seconds per clip.
• Output resolution is limited to 480P or 768P.
• If you supply a first frame image, aspect ratio controls won’t override the image’s shape.
• Fine details can drift across frames. Small logos, hands, and faces may change.
• The model can follow the “vibe” of a prompt while missing key constraints. Review each take closely.
• Low-quality inputs reduce control. Blurry, heavily compressed, or low-resolution first/last frames can cause unstable motion and artifacts.

Safety & compliance

Follow MiniMax’s acceptable use rules when generating video content. Don’t create illegal content or content that violates privacy, consent, or intellectual property rights. If you generate content depicting real people, get permission and avoid deceptive or harmful deepfakes.