H3 Max Image to Video Generator
Use a source image to anchor composition and appearance, then direct subject motion, camera movement, timing, and sound.
Start with a clean source frame
Choose an image with a readable subject, intentional composition, and enough space for the movement you want. The source already defines appearance, so the prompt should concentrate on what changes after the first frame.
Choose Turbo or H3 Max
The Turbo image-to-video endpoint is designed for quick iteration. The H3 Max image route supports a start image and can also use an end image when the final composition matters. Available controls depend on the selected route and remain visible in the generator.
Describe motion rather than restating the image
Specify who moves first, how the camera responds, what remains stable, and how the shot ends. Mention hair, fabric, reflections, smoke, water, or other secondary motion only when it supports the main action.
Control cost and iteration
Credits are based on model, duration, and resolution. Test uncertain movement at a shorter duration, preserve the source frame between versions, and revise one motion instruction at a time so differences are easy to judge.
Questions
Frequently asked questions
Can I provide both a start and an end image?
The H3 Max image workflow can expose an optional end frame. Turbo image to video is intended for a simpler start-image workflow.
Should the prompt describe everything visible?
No. Use the image for appearance and composition, then spend prompt space on motion, camera behavior, timing, sound, and constraints.
Why does the subject change?
Ambiguous motion, occlusion, extreme camera moves, or conflicting end-frame details can reduce consistency. Simplify the action and state what must remain unchanged.