NewNano Banana 2.1 is liveTry it

Vidu Q4 AI Video Generator

Vidu Q4 online: Shengshu's new Preview model turns a photo or up to 8 reference images into a clip with sound, 5–16 seconds.

Start frame
Prompt0/20000
Quality
Length5s
Generate audio

One video per run.

Required credits200
Credits by planUnlimited: 120
Free
—
Starter
160
200
Pro
140
200
Unlimited
120
200

Sample videos

Vidu Q4 is Shengshu Technology's next flagship video model; the version available now is the Preview, released on October 7, 2026. It animates a start image, keeping the photo's shape, or builds a scene from up to 8 reference images with up to three MP3 voice or sound clips, and generates synced audio and lip movement. On Vidlark it runs at 540p, 720p, 1080p, 5–16 seconds, from 20 credits a second, and it is available on any plan or credit pack.

Updated

At a glance

Made by
Shengshu
Price
From 20 credits per second
Quality
540p, 720p, 1080p
Length
5s, 8s, 10s, 16s
Input
A photo or reference images
Aspect ratio
16:9, 9:16, 1:1, 4:3, 3:4
References
Up to 8 images
Audio references
Up to 3 MP3 clips
Sound
Yes, generated with the video
Per run
One video
Available on
Any plan or credit pack
Account
Free account required, 100 credits on sign-up, no card
Unlimited plan
40% off

Vidlark is an independent platform and isn't affiliated with Shengshu.

How to use the Vidu Q4 AI video generator

  1. 1

    Describe the shot

    Type a prompt, or upload a photo to animate.

  2. 2

    Pick the settings

    540p / 720p / 1080p, 5–16s; the price shows before you start.

  3. 3

    Generate and download

    Download the video when it is ready.

What it does well

Faces that act

Shengshu tuned Q4 for expression, emotion and body movement, so a smile, a glance or a shrug reads clearly instead of drifting. It suits close-ups and short scenes built around a person.

One cast from many references

Upload up to 8 images of the people, the place and the style, then refer to them by order in the prompt ("the man in image 2"). They stay recognisable across the clip.

Voices with lip sync

Add up to three short MP3 voice clips and say who speaks; the lines come out in that voice with matching mouth movement. Without clips, sound is generated from the prompt, or switched off.

Settings and prices

Video is billed per second of output; the button always shows the exact total before you generate.

SettingPriceExampleUnlimited plan
540p20 credits / second5-second clip = 100 creditsUses credits
720p40 credits / second5-second clip = 200 creditsUses credits
1080p50 credits / second5-second clip = 250 creditsUses credits

Prompt tips for better videos

  • Number your references

    The model reads references by order. Name them that way in the prompt: "the woman in image 1 hands the cup from image 3 to the man in image 2".

  • Draft at 540p

    Settle the action at 20 credits a second, then render the keeper at 1080p for 50 a second.

  • Use MP3 for voices

    Vidu Q4 takes reference audio as MP3 only, 3 to 12 seconds a clip. Trim a clean line of speech without music under it.

Example prompt

A cyclist rides through a rainy neon street at night, puddles splashing under the wheels. Slow tracking shot from the side, 35mm, reflections on the wet asphalt. Sound: rain, tires on water, distant traffic.

Use this prompt

What people make with it

01

Short drama scenes

A two-person exchange with a reaction shot, from character photos and a line of dialogue, in up to 16 seconds.

02

Product spots with a presenter

Combine the product shot, a presenter photo and the setting as references for a short ad where the same face and pack appear throughout.

03

Bringing a portrait to life

Start from one photo and add a small action and a line of speech; the clip keeps the photo’s framing and look.

How it compares

How it lines up against similar video models on this site; every one uses the same credit balance.

ModelBest forTop qualityLengthSoundFrom
Vidu Q4Character scenes with voices from photos and references.1080p5–16sYes20 credits/s
Kling 3.0Realistic motion and camera moves, with sound when needed.1080p5–15sYes35 credits/s
Seedance 2.0Multi-shot clips with steady characters, up to 1080p.1080p5–15sYes25 credits/s

Other models to try

All models

Vidu Q4 questions

Is there a free website to try Vidu Q4?

Vidu's own site and app run it on their credits. On Vidlark it is available on any plan or credit pack, so the 100 sign-up credits do not cover it; the smallest run, 5 seconds at 540p, is 100 credits and nothing is charged if a run fails.

What is Vidu Q4?

The next flagship video model from Shengshu Technology, the company behind Vidu. What is out now is Vidu Q4 Preview, released on October 7, 2026 ahead of the full model: it turns a photo, or a set of reference images with optional voice clips, into a clip with sound, and it is tuned for faces, emotion and body movement.

Can Vidu Q4 make a video from text alone?

Not yet. The Preview takes a start image or reference images; prompt-only text to video is not part of this release. For a clip from words only, Kling 3.0 or Seedance 2.0 on this site take a prompt alone.

How much does Vidu Q4 cost?

It is billed per second: 20 credits at 540p, 40 at 720p and 50 at 1080p, with sound or without at the same price. A 10-second 720p clip is 400 credits; plans take a discount off that.

How long can the videos be?

5, 8, 10, 16 seconds, at 540p, 720p, 1080p. Each run makes one video; for longer stories, generate several shots and edit them together.

Credits, refunds, watermarks and commercial use work the same on every model. See the pricing FAQ

Make a Vidu Q4 video

Pick a plan or a credit pack and start. Failed runs are refunded.

Generate a video