Choose a category
3 models available
8 models available
3 models available
5 models available
4 models available
8 models available
Animate a first frame, work toward a last frame, or connect both with sound.
Add a first frame, a last frame, or both. JPEG, PNG or WebP, up to 30MB each.
Describe the motion between frames and the sound. Keep the subject and composition consistent.
Choose 4–15 seconds. Longer clips use more credits.
Start at 768P; choose 2K for a higher-resolution output.
Give MiniMax H3 a first frame, a last frame, or both. Describe the motion and sound to create a 4–15 second video at 768P or 2K. Use one image for a starting composition, one closing image for an intended destination, or two frames for a planned transition.
Use frames that belong to the same scene and explain how the camera travels between them.
Examples for 5 seconds with 1 frame, including generated audio. The current price depends on duration and resolution; these image inputs have no separate surcharge.
| Resolution | Duration | Flow credits |
|---|---|---|
| 768P | 5s | 9 |
| 2K | 5s | 15 |
Credits are held on submission and released for confirmed failed generations. Compare credit plans and other models.
Choose the controls that fit your source material.
The first-frame slot gives the shot an opening visual reference. Describe movement instead of listing everything already visible: a curtain lifting, a product rotating or the camera moving closer. Small, specific actions are easier to evaluate than many unrelated changes.
A last frame is also accepted on its own. Tell the model what should lead into that image. When both frames are supplied, describe a plausible transition between them; very different compositions may require a bigger change than the clip can express cleanly.
This workflow has no aspect-ratio selector. Prepare your images in the composition you intend to use. For two-frame shots, matching dimensions and a consistent subject make the intended continuity clearer.
Ask for ambience, effects or dialogue alongside the visible action. H3 generates an audio track with the video. The frame uploads do not provide recorded audio, and this page does not accept a reference video or audio file.
Change one part of the input at a time so you can judge what helped.
Prepare the input, describe the change, then review the credit cost.
Add a JPEG, PNG or WebP image up to 30MB to either slot. At least one frame is required. Use the first slot for an opening composition and the last slot for the intended ending.
Try “Slowly turn the ceramic vase toward the camera while the afternoon light shifts across it. End on the closer view. Quiet room ambience.” Avoid asking for an unrelated location or a new subject unless that change is intentional.
Start at 768P and 5 seconds, or select another duration up to 15 seconds. The current price depends on duration and resolution. Review the final result for continuity before making another attempt at 2K.
Change one part of the input at a time so you can judge what helped.
An original H3 result generated on Flow Veo 3 from the inputs shown below. The inputs are frames from our H3 greenhouse text-to-video example. Play to hear the generated audio.
Prompt: The orange butterfly slowly opens its wings and lifts off the green leaf into warm sunrise. Move smoothly from the opening frame to the closing frame. A gentle camera push, quiet birdsong and soft rustling leaves. No music, text or logos.
Answers about this workflow, inputs and output settings.
No. You may upload only a first frame, only a last frame, or both. At least one is required. With no images, use the H3 Text to Video page.
It guides the intended ending of the generated shot. Describe the action that connects the opening and ending, or the action that should arrive at a closing image when only the last frame is supplied. Results can vary; it is not a guarantee of pixel-perfect reconstruction.
The H3 image-to-video interface takes its framing from the supplied images and has no separate aspect-ratio parameter. Compose your source images for the destination before uploading.
No. Under the current image-only pricing, one or two frames cost the same at a given duration and resolution. The exact credit total is shown before submission.
Frame inputs describe the opening or ending composition. Reference images provide visual material for subjects and scenes without choosing a fixed opening or closing frame. Use Reference to Video if you want to assign several images separate roles.
JPEG, PNG and WebP, up to 30MB per image. Choose clear source images that you have permission to use. Video clips and audio uploads are not supported in this workflow.