Animate a still life
Start with a product or object photograph and ask for a gentle camera movement.
Image and reference video
Explore Vidu Q4 Preview with one starting image or a prompt and optional references. Choose the motion, video length, and resolution for your next short clip.
Video generation is currently unavailable on this site.
Vidu Q4 turns a visual idea into a short video workflow. Start with a still image when the composition already exists, or use reference mode to describe a scene and guide it with optional pictures. In either case, a useful prompt focuses on what moves and how the camera sees it.
The version described here is Vidu Q4 Preview. This is an independent service, with two modes rather than a separate tool for every type of input. Read how to use Vidu Q4 for the individual controls, upload steps, and ways to prepare a clear first prompt.
Choose the mode that matches your starting material. References provide guidance; they do not guarantee identical subjects or perfect motion in every frame.
| Control | Image to Video | Reference to Video |
|---|---|---|
| Starting input | One image; prompt optional | Required prompt; images optional |
| Reference material | One starting picture | Up to 12 pictures and 3 audio references |
| Frame shape | Based on your starting image | 16:9, 9:16, 1:1, 3:4, or 4:3 |
| Prompt expansion | Available, on by default | Not available in this mode |
| Video settings | 540p to 4K; 3 to 16 seconds | 540p to 4K; 3 to 16 seconds |
Use one image to establish the subject, lighting, and composition. Describe a small movement first: a slow camera push, drifting clouds, or a gentle turn. A narrow motion goal makes it easier to judge the result.
| Setting | How to use it |
|---|---|
| Image | Upload a PNG, JPG, JPEG, or WebP file, up to 100MB. |
| Prompt | Optional. Describe movement and camera direction rather than restating every visible detail. |
| Resolution and duration | Choose an output size and a whole-number length from 3 to 16 seconds. |
| Prompt expansion | Allow extra interpretation, or turn it off to try your own precise wording. |
| Generate audio | Turn generated audio on or off for the clip. |
| Seed | Leave blank for a random seed, or enter an integer to try a specific value. |
Write a required prompt and add relevant reference pictures when you need visual guidance. Explain which subject or setting matters, then describe the action. You can also work from the prompt alone without adding images.
| Setting | How to use it |
|---|---|
| Prompt | Describe the subject, setting, action, and camera. This field is required. |
| Reference images | Optional. Use up to 12 pictures that support the same scene or subject. |
| Reference audio | Optional, up to 3 references. This feature is not voice cloning. |
| Aspect ratio | Choose landscape, portrait, square, 3:4, or 4:3 for the intended screen. |
| Resolution and duration | Choose the same video sizes and lengths as image mode. |
| Seed | Leave blank or use 0 for random; use a nonzero integer for a specific value. |
Match the clip to its destination before increasing every setting. Longer clips allow more action, while higher resolutions use more credits. Compare Vidu Q4 pricing when choosing a length and size.
| Choice | Useful starting point |
|---|---|
| Quick exploration | Try a short, lower-resolution clip to check one movement. |
| Default settings | 720p and 5 seconds offer a straightforward starting point. |
| More detail | Consider 1080p, 2K, or 4K when your intended use needs a larger frame. |
| More time | Use 3 to 16 seconds, with enough time for the action you describe. |
| Audio | Choose whether the video should include generated audio. |
Once generation is available, follow one mode from input to result. Keep your first attempt focused on a single scene rather than several unrelated actions.
Start with an idea that fits a short clip. Browse Vidu Q4 prompt ideas for suggested descriptions you can adapt to your own material.
Start with a product or object photograph and ask for a gentle camera movement.
Choose references that show a subject clearly, then describe one action.
Use reference mode with a prompt alone to describe a landscape, object, or simple scene.
Choose a portrait or landscape frame in reference mode.
Use your own or authorized material and review each output before sharing it. Read the terms of service for account and content rules.
Image to Video requires one picture in PNG, JPG, JPEG, or WebP format. Reference to Video requires a written prompt; pictures and audio references are optional. Choose material you have permission to use and keep each uploaded file within 100MB.
Yes. Write a prompt describing the subject, setting, and movement, then leave the reference image list empty. You are still using Reference to Video. Add pictures when visual guidance is useful, rather than uploading unrelated images just to fill the list.
Both modes offer 540p, 720p, 1080p, 2K, and 4K, with whole-number durations from 3 to 16 seconds. The starting settings are 720p and 5 seconds. Higher settings are available options, rather than a guarantee of a successful or flawless result.
In Image to Video, prompt expansion lets the model elaborate on your description. Leave it on for a simple direction, or turn it off when you want to try a more specific prompt. This switch is separate from Generate audio and is not present in reference mode.
Vidu Q4 uses a per-second credit rate for the selected resolution, multiplied by the video length. Both modes use the same rates. Review the estimate before submitting and choose a shorter clip when you want to explore an idea with fewer credits.