Vidu Q4 · Reference to Video
Vidu Q4 Reference to Video Generator
Feed Vidu Q4 up to 12 reference images (characters, objects, scenes) plus optional voice references, and generate shots where identity stays stable. Five aspect ratios, 3-16 seconds, up to 4K with native audio.
Why reference-to-video changes the workflow
Multi-image conditioning is the signature Vidu Q4 strength: define who and what appears once, then direct new shots by prompt.
Up to 12 reference images
Mix character sheets, props, and locations. Mention them in the prompt with @image-1 style tags.
Voice references
Attach up to 3 audio or voice clips so generated speech keeps a consistent voice.
Five aspect ratios
Pick 16:9, 9:16, 1:1, 4:3, or 3:4 for the target platform instead of inheriting a source frame.
Native audio + 4K
Synchronized sound and lip-matched dialogue in the same pass, delivered at up to 4K.
How it works
From references to a consistent shot in three steps.
Upload references
Add 1-12 images of your character, product, or scene. Optionally attach voice or audio references.
Tag them in the prompt
Write the shot and cite references with @image-1, @image-2, etc. Set aspect ratio, duration (3-16s), and resolution up to 4K.
Generate and iterate
Submit, review the result in the workbench, then adjust references or prompt for the next shot.
Reference-to-video vs image-to-video
Both Vidu Q4 modes start from images. Reference mode adds multi-asset consistency control.
| Reference to Video | Image to Video | ||
|---|---|---|---|
| Inputs | 1-12 images + up to 3 audio refs | 1 start image | |
| Aspect ratio | Choose 16:9, 9:16, 1:1, 4:3, 3:4 | Follows the source frame | |
| Consistency goal | Same character/product across new shots | Animate what is in the frame | |
| Best for | Series, ads, recurring characters | Quick one-off animations |
Both modes produce 3-16 second clips at up to 4K with optional native audio.
What creators use it for
Recurring characters
Keep a mascot or hero character identical across an episode or campaign.
Product ads
Lock a product's look from reference shots, then place it in new scenes by prompt.
Storyboards to shots
Turn character sheets and set references into moving shots without reshooting.
Voice-led scenes
Attach a voice reference so the generated dialogue sounds like the same speaker.
Limits to know
Honest constraints from the published Vidu Q4 Preview spec.
12 images, 3 audio clips
This site currently accepts up to 12 reference images and 3 audio references per generation.
Prompt tags matter
Cite each reference explicitly (@image-1, @image-2). Untagged references may be ignored.
Clip length 3-16s
Shots are short by design. Build sequences by chaining clips in an editor.
Credits-based pricing
Generation cost scales with duration and resolution. The workbench shows the exact credit cost before you submit. Plans and credit packs are on the pricing page.
Reference-to-video FAQ
How many reference images can I upload?
Up to 12 reference images per generation on this site, and you can mix characters, objects, and scene references.
How do I tell Vidu Q4 which image is which?
Mention them in the prompt with tags like @image-1 and @image-2 in upload order. Explicit tags keep assignments unambiguous.
Can I add a voice or music reference?
Yes, up to 3 audio references are supported, useful for keeping a character's voice consistent.
Which aspect ratios are available?
Reference-to-video lets you pick 16:9, 9:16, 1:1, 4:3, 3:4, unlike image-to-video, which follows the source frame.
Is this the official Vidu site?
No. vidu-q4.com is an independent third-party workbench for Vidu Q4 Preview and is not affiliated with Vidu or ShengShu Technology.
Keep your characters consistent
Upload references, tag them in the prompt, and direct the next shot.