Generate Video in Up to 2K at 24fps
Choose 768P for drafts or 2560×1440 output at 24 frames per second when the final clip needs more visible detail.
Create videos from text, frames, and references
Create 5-15 second videos from text, frames, and references with synchronized audio and output up to 2K.
Create with MiniMax H3
Generate 5-15 second videos from text, a required first frame with an optional last frame, or multimodal references. Choose 768P or 2K output and customize your generation settings.
Reference Media
Max 12 total: 9 images, 3 videos, 3 audio. Media 2-15s. Files 10/50/10MB.

1/12 files, 1/9 images, 0/3 videos, 0/3 audio
Available credits: 0
BIWA Character Introduction
Create a 15-second premium cinematic character reveal using the BIWA character sheet as the sole visual authority.
MiniMax H3 accepts text, image, video, and audio references. Combine only the materials your scene needs for more precise creative direction.
Prompt Guide
Name uploaded files directly in your prompt.
Use references such as @Image 1, @Video 1, or @Audio 1 so the model knows which material to use.
Material Limits
Supported Input Combinations
Note: Reference images, videos, and audio are optional. Add only the materials needed for your prompt.
Images
JPG, JPEG, PNG, WEBP, BMP, TIFF, or GIF. Up to 10 MB each.
Videos
MP4 or MOV. 2 to 15 seconds, up to 50 MB each.
Audio
MP3 or WAV. 2 to 15 seconds, up to 10 MB each.
Independent third-party service. Not affiliated with or operated by MiniMax.
Model specifications checked against MiniMax's published H3 documentation. The hosted interface may expose a narrower range than the underlying model. Last verified September 10, 2026.
See the model in motion
Watch short videos created for product advertising, cinematic storytelling, fashion, music, game concepts, and interface animation. Hover over a video to play it with sound.
Control the image, motion, and sound
Create short-form AI video with 768P or 2K output, synchronized stereo audio, first-and-last-frame control, mixed references, and six aspect ratios.
Choose 768P for drafts or 2560×1440 output at 24 frames per second when the final clip needs more visible detail.
Generate dialogue, ambience, music, and physical sound in the same pass so audio follows the action on screen.
Combine up to nine images, three video clips, and three audio files to guide identity, movement, camera language, style, and sound.
Request a focused change to the action, framing, lighting, or sound while preserving the parts that already work.
Create title sequences, packaging shots, brand graphics, subtitles, and animated UI that remain part of the scene.
Generate landscape, square, portrait, or cinematic video in 21:9, 16:9, 4:3, 1:1, 3:4, or 9:16.
The model behind the generator
MiniMax H3 is a multimodal AI video model that creates short clips from written prompts, frames, and mixed media references.
MiniMax H3 understands text, images, video, and audio together, then generates the moving picture and stereo soundtrack as one result.
Use a written prompt when you want the model to invent the scene. Add a first or last frame when composition matters. Supply reference media when a subject, product, movement, camera rhythm, visual style, or sound needs clearer direction.
A single request can combine up to 12 reference materials: as many as nine images, three video clips, or three audio files, plus a prompt of up to 7,000 characters. Results still depend on input quality and the complexity of the requested action.
For practical prompt structures and reference instructions, read the MiniMax H3 prompt guide.
2K
Native resolution
24fps
Playback
12
Mixed references
7,000
Prompt characters
Stereo
Audio in one pass
From brief to finished clip
Choose the right generation mode, describe the shot and sound, add only the references you need, then select the output settings.

Start with text-to-video, first-and-last-frame control, or multimodal reference video. Describe the subject, action, camera, lighting, pacing, and sound.

Attach only the images, clips, or audio that have a clear role, then choose an aspect ratio, 768P or 2K resolution, and a 5–15 second duration.

Watch the complete clip with sound. If one element misses the brief, request a specific change instead of replacing every successful detail.
Choose the right input
Start with the material you already have. Each MiniMax H3 workflow offers a different balance between invention, composition control, and reference fidelity.
From concept tests to publishable shots
Use MiniMax H3 for short-form social content, product advertising, ecommerce concepts, game and interface motion, or film previsualization.
Content Creators
A vertical clip comes back with sound already attached, ready for the next publishing step.
Marketers and Agencies
Compare campaign directions before committing a production budget.
Ecommerce Sellers
Put the product images you already own into motion at catalog scale.
Game and App Teams
Create character reels, CG concepts, and interface motion without a long contractor cycle.
Film and Animation Studios
Use real camera vocabulary to turn a storyboard note into a reviewable shot.
Plans and generation costs
Choose a one-time credit pack based on the number, duration, and resolution of the videos you plan to generate. Purchased credits do not expire.
Free
One 5-second 2K clip on signup, with no card and no stripped-down model.
Starter
For testing prompts, references, aspect ratios, and a few proper re-renders.
Basic
For regular publishing, longer clips, variations, and stacked reference files.
Professional
For production sprints, ad campaigns, client work, and teams generating at volume.
Frequently asked questions
Clear answers about text-to-video, frame control, multimodal references, resolution, clip length, synchronized audio, and credit usage.
New accounts can start with 20 credits, enough for one 5-second 2K clip.
Choose a workflow, enter a prompt or upload references, select the ratio and duration, then generate.
Yes. Describe the subject, action, camera, lighting, and sound in a prompt up to 7,000 characters.
Yes. The image defines the look while the prompt defines the movement and sound.
Yes. Upload opening and closing images and let H3 generate the motion between them.
Yes. H3 accepts up to three audio references to guide timing, pacing, and atmosphere.
H3 supports 5 to 15 second clips at up to native 2K and 24fps.
Yes. Dialogue, ambience, music, and effects can be generated in the same pass.
A request can contain up to nine images, three video clips, and three audio tracks.
Output is charged by duration and resolution. A 5-second 2K clip uses 20 credits.
No. MiniMaxm is an independent third-party experience and is not operated by MiniMax.
Start with a prompt, one or two frames, or a set of image, video, and audio references. Generate a 5–15 second clip with synchronized stereo sound in 768P or 2K.
Open the MiniMax H3 Generator