Vidu Q3 AI Video Generator

Create a short video with Vidu Q3. Start with a prompt or image and direct the motion, framing and optional sound.

Try an idea

Make the starting frame do some of the work

Vidu Q3 supports both text and image starts in this workflow. An image can establish the subject’s appearance while your prompt explains what moves. That separation is useful when you already like a composition and want to explore a small action inside it.

Only one starting image is accepted here. End-frame and multiple-reference controls are not exposed, so avoid describing them as fixed constraints on the result.

Vidu Q3 prompt examples

A fuller brief for the moment when you know what you want. Copy it, or send it to the creator and make it your own.

Text to video · one continuous shot

Direct the whole shot

An illustrated girl watching lanterns drift over a lake. Lanterns move slowly upward while the adult character turns her head slightly. Keep the camera steady.

Establish the subject clearly in the opening frame. Keep its appearance, proportions and direction of light consistent throughout. Let the action unfold at a measured pace, with no abrupt camera changes or transitions to another location.

Use a coherent foreground, middle distance and background. End with the main subject still clearly visible. No cuts, captions or added lettering. Sound direction: quiet environmental ambience appropriate to the setting, with no music or spoken dialogue.

Image to video · add your own image first

Animate a starting image

Use the uploaded image as the starting composition. Preserve its subject, colors, material appearance and overall lighting. Lanterns move slowly upward while the adult character turns her head slightly. Keep the camera steady.

Keep secondary movement restrained so the main action remains readable. Do not introduce new people, change the setting or redesign the subject. Maintain a steady pace and finish with a composition close to the starting view.

One continuous shot with no cuts or captions. Avoid sudden zooms or rapid camera rotation.

How to get started with Vidu Q3

Set the scene

Begin your Vidu Q3 prompt with the subject and location. If you upload an image, use it to establish the appearance and spend the prompt on the movement you want.

Direct one action

Name what moves and how the camera behaves. “The cyclist passes from left to right; the camera stays still” is clearer than several conflicting camera directions.

Match the pace to the length

Give a short clip one readable gesture. For a longer clip, describe a gradual movement with an intentional endpoint. Check the selected duration again after changing resolution.

Vidu Q3 settings on Maxsa

Starting point	Text prompt or one reference image
Framing	16:9 / 9:16 / 1:1
Resolution	360p / 540p / 720p / 1080p
Length	5 / 10 / 15 seconds
Audio	Optional audio

These are the options connected to this creator. The model may offer additional features in other products. The credit quote is shown before you confirm generation.

FAQs about Vidu Q3

What is Vidu Q3?

Can I start with an image in Vidu Q3?

Yes. Upload one JPG, PNG or WebP image under 10 MB. Use the image to establish the starting frame, then describe the action and camera movement. This creator does not accept multiple reference images.

Does Vidu Q3 generate sound?

Yes, you can request audio using the Audio control. Describe the ambience or sound effect in your prompt, then review the finished clip with sound on.

Which clip lengths does Vidu Q3 support?

The current length options are 5 / 10 / 15 seconds. Choose one action that fits that time rather than asking for several scenes.

How is Vidu Q3 different from the other versions?

This is the ShengShu video version currently listed here. The links below offer other model families if you want to compare a different approach. Only one starting image is accepted here. End-frame and multiple-reference controls are not exposed, so avoid describing them as fixed constraints on the result.

Where can I find and save my result?

The finished video appears in Your creations below the generator, where you can preview and download it. Save a copy you want to keep. If a result is stored only on this device, the result panel labels it accordingly.