Minimax H3 video generator
- Type
- Video model
- Made by
- MiniMax
- Input
- Text, plus reference images and sounds, or a first and last frame
- Output
- MP4 video
- Options
- 6 aspect ratios · Duration 4–15 s · Resolution: 768p, 2K · First & last frame
- Price
- From 1,000 credits
- Plan
- Paid plans
Minimax H3 is MiniMax's video model for 2K footage up to 15 seconds long. Describe a shot, optionally attach references or a start and end frame, and it renders an MP4 at 768p or 2K in one of six aspect ratios.
What you can do with it
- Text to video. Write a scene and get a clip of 4 to 15 seconds.
- References. Attach images and sound clips of people, products or places for the clip to work from. Up to 10 files per run.
- First and last frame. Switch on First & last frame, add the opening frame (required) and an optional closing frame, and the clip travels from one to the other.
How to use it
- Write the shot in the composer.
- Attach references, or switch on First & last frame and add your frames.
- Set aspect ratio, duration and resolution with the chips.
- Check the price in the composer, then generate.
Writing prompts for Minimax H3
- Lead with the subject and the action. "A cyclist crests a hill at sunrise" gives the model something to animate right away.
- Then set the scene and the look. Location, time of day, weather, lens feel, color palette.
- Give the camera one job. A slow pan, a tracking shot alongside the subject, or a locked-off frame. One clear move usually reads better than several.
- Keep it to one continuous moment. Fifteen seconds is long enough for an action to unfold, but a list of unrelated events tends to blur.
- With a first frame, describe the motion and camera rather than what is already in the image.
Example prompts:
- A fishing boat leaves a misty harbor at dawn, gulls circling the mast. Wide shot from the pier, slow pan following the boat out to sea. Cool blue light, soft film grain.
- Close-up of hands kneading dough on a floured wooden counter, then pulling back to show a small bakery kitchen. Warm morning light through the window.
Choosing settings
- Aspect ratio: 16:9 (default), 9:16, 4:3, 3:4, 1:1 or 21:9. Use 9:16 for vertical social video and 21:9 for a wide, cinematic frame.
- Duration: 4 to 15 seconds, default 8.
- Resolution: 768p (default) or 2K.
- First & last frame: off by default.
Pricing is per second of video: 250 credits at 768p and 350 credits at 2K. Multiply by the duration to get the cost of a clip: a 4-second 768p clip is 1,000 credits, an 8-second 2K clip is 2,800, and a 15-second 2K clip is 5,250. Drafting at 768p and re-running the keeper at 2K is a simple way to save credits. The composer shows the exact price before you generate.
If you need sound generated with the picture, look at Seedance 2.0 or Wan 3.0, which have a Sound switch.