MiniMax · Text to video · Image to video · Reference images to video
MiniMax H3 AI Video Generator — 15s Video With Native Audio
Create up to 15-second MiniMax H3 videos from text, a photo or multiple references — images, a motion video and audio — with native stereo sound.
What MiniMax H3 is good at
- Up to 15 seconds
- Native stereo audio
- Up to 9 image + 3 video + 3 audio references
- 480p and 768p, 6 aspect ratios
How to use MiniMax H3
Pick the model
Open the generator and pick MiniMax H3.
Add your idea
Write a prompt, or switch to reference mode and add photos or a saved character.
Choose settings
Choose duration, resolution, aspect ratio, then press Generate.
Download
Download the result or find it later in My Creations.
Priser för MiniMax H3
Krediter per video. Misslyckade eller blockerade genereringar återbetalas automatiskt.
| Upplösning | Krediter | |
|---|---|---|
| 5s | 480P | 12 |
| 10s | 480P | 23 |
| 15s | 480P | 34 |
| 5s | 768P | 23 |
| 10s | 768P | 45 |
| 15s | 768P | 67 |
Vanliga frågor
How much does MiniMax H3 cost on Vadu?
MiniMax H3 starts at 12 credits per video. The exact cost depends on duration and resolution and is shown on the Generate button before you start. Failed generations are refunded automatically.
What can I create with MiniMax H3?
MiniMax H3 supports text to video, image to video, reference images to video, 5 / 10 / 15 second clips and 480P / 768P output. In reference mode you can add up to 9 reference images to keep the same person, outfit or product consistent.
Can I reuse my photos and characters?
Yes. Everything you upload is saved to your library, and you can save a person as a character (1–4 photos). Pick them from the library in any Vadu tool instead of uploading again.
What content is not allowed?
Prompts, uploads and results are checked automatically. Nudity, sexual content, content involving minors in any sexual or intimate context, graphic violence and deceptive use of real people are blocked. Blocked generations are refunded.
Andra videomodeller
Fast Veo 3.1 with native audio — the best value for realistic clips
Google’s flagship video model with reference images and native audio
Realistic motion and strong prompt adherence, 5–10s with audio
Multi-reference video: keep up to 7 people, outfits or products consistent
Synced audio-video with lip-sync, up to 9 reference images
Stable characters, start/end frames, reference images, videos and audio
Expressive, stable motion across art styles
Cheapest way to animate a photo — great for quick tests