Text to video · Media to video
MiniMax H3
Arrange text, images and audiovisual references into 4–15 second clips at 768P or 2K. Choose image guidance or multimodal references, with sound generated alongside the video.
- Video duration
- 4–15 seconds
- Formats and limits
- 2K · 768P
- Reference media
- Images 9 · Video 3 · Audio 3
OpenAI's latest video generation model with enhanced quality
Google's latest model with exceptional quality output
Seedance 2.5: text, frame and media reference video generation
With one image, Grok follows the image aspect ratio.
Seedance and MiniMax also bill the actual duration of input videos. Documents: maximum 50 pages.
When this option is enabled, the output video may be selected by AI Generate and published to the Explore.
Choose your settings and generate. Preview and download the result when it is ready.
MiniMax H3 · Features & workflow
Match the mode to your material
Start with text for a written idea, image mode for a starting still, or reference mode when several visual and audio materials need to work together.
Include sound in the scene
Describe dialogue, music and ambience in the prompt. Pair an audio reference with an image or video so the soundtrack has a visual context.
Formats & credits
| Formats and limits | credits / second |
|---|---|
| 768P | 8 |
| 2K | 13 |
Input and output video durations are both billed. The total is rounded up to a whole credit.
The first 5 reference images are free; each additional image costs 4 credits. Audio input is free.
MiniMax H3 · Questions, answered
What duration and resolutions can MiniMax H3 generate?
Choose 4–15 seconds at 2K · 768P. Plan the shot’s action first, then its duration. Resolution and aspect ratio are separate settings.
Which references can I use with MiniMax H3?
Use up to 9 images, 3 videos and 3 audio clips. Reference video and audio each have a separate combined limit of 15 seconds. Explain which materials should guide appearance, action or sound.
Can I upload only audio to MiniMax H3?
No. MiniMax H3 reference mode needs at least an image or video alongside audio. It accepts up to 9 images, 3 videos and 3 audio clips, with a 15-second total for reference video and a separate 15-second total for audio.
How do I start creating with MiniMax H3?
Start with text for a written idea, image mode for a starting still, or reference mode when several visual and audio materials need to work together.
How does MiniMax H3 handle sound?
MiniMax H3 generates sound with the video; this workspace has no audio-off option. The first 5 reference images add no image fee; each further image adds 4 credits. Audio input adds no fee.
How are MiniMax H3 video credits calculated?
768P: 8 credits / second; 2K: 13 credits / second. Input and output video durations are both billed. The total is rounded up to a whole credit. For example, 768P: (5 + 5) × 8 = 80.