Vidu Q4

Vidu's newest video model: up to 12 reference subjects and 3 reference voices, 3 to 16 second clips, native audio, up to 4K.

Ingen abonen

Krediter aldri eksisterer

Lær mer

Betale en gang for kreditter - bruke dem på alle modeller på ZOOOP. · Fyll på når du trenger det, ingen månedlig avgift.

Vidu Q4

Velg Modell*

Prompt*

0/5000

Reference Images*

Reference Voices

Spektforhold*

Oppløsning*

Gtid*

Generate Audio

Det du genererer, vises her når du logger inn.

Gratis kreditter når du registrerer deg
Ingen abonnement · kredittene utløper aldri
Hver topp AI-bilde, video og lydmodell
Lagre, organiser og last ned alle verkene dine

Det er gratis å opprette en konto.

Nøkkelegenskaper

Up to 12 reference subjects

Attach up to 12 images of characters, products or places and the model keeps each one consistent through the clip.

Reference voices

Add up to three MP3 voice clips of 3 to 12 seconds, and the generated dialogue follows those voices.

540p to 4K

Five resolution tiers: draft at 540p, then render the chosen take at 1080p, 2K or 4K for delivery.

3 to 16 seconds with sound

Any length from 3 to 16 seconds, with synchronized dialogue, effects and ambience. Turn audio off when you only want picture.

Bruk saker

Brand characters in new scenes

Keep a mascot or spokesperson identical while placing them in a kitchen, a street or a stage.

Product ads at 4K

Animate a product still into a polished shot and render the final at 4K for large screens.

Consistent voices for a series

Reuse the same voice clip across episodes so a narrator or character sounds the same every time.

Multi-character scenes

Combine several character references in one shot, such as two friends meeting in a café.

Bring a photo to life

Start from a single photo and let the scene move: hair in the wind, a smile, a camera drift.

Vertical clips for social

Render reference scenes in 9:16 for vertical short videos, or 1:1 and 4:3 for feeds.

Velg riktig modell

Vidu Q4 is the reference specialist: many subjects, consistent voices, up to 4K. Switch when your shot needs something else.

Many consistent subjects and voices, up to 4KVidu Q4 ←

Start-and-end-frame transitionsMiniMax H3 Max

Text-to-video, plus edit and extendGemini Omni 1.1 Flash

Swap a person into an existing clipMiniMax H3

Takes up to 30 secondsWan V3.0

The previous Vidu generationVidu Q3 Pro

Hvordan bruker

Open Vidu Q4 from this page and pick a mode: start frame, or reference images (with optional voices).

Upload your start frame or reference images, and write what should happen.

Choose duration (3 to 16 seconds), resolution (540p to 4K) and, for references, the aspect ratio.

Generate at a low tier to check the take, then re-render at a higher resolution.

Deep dive

What Vidu Q4 is good at, and where it isn't

Vidu Q4 is built for keeping many things consistent at once. Where most video models are comfortable with one or two references, Q4 takes up to 12 reference images in a single shot and up to three reference voices, so a cast of characters, a product and a location can all hold their look while the generated dialogue keeps the right voice.

The first flagship capability is reference-to-video with subjects and voices. Attach the images, tag who does what in the prompt, and optionally add MP3 voice clips of 3 to 12 seconds. In our tests two robot references stood side by side on a sunset rooftop in 9:16, swaying without their designs swapping, and a single character waved hello to the camera with sound generated alongside the picture.

The second is resolution range. Five tiers from 540p to 4K mean you can test a take cheaply and render the keeper at 2K or 4K for large screens, without switching models.

The third is length with sound. Any duration from 3 to 16 seconds, with dialogue, effects and ambience generated together. You can switch audio off for picture-only clips.

Where it's weaker: there is no text-only mode and no end frame. Pure text-to-video belongs to Gemini Omni 1.1 Flash or MiniMax H3 Max, and start-and-end transitions to MiniMax H3 Max. It is also still a preview model, so expect behavior to change as it matures. For takes longer than 16 seconds, Wan V3.0 goes to 30.

A reasonable mental model: choose Vidu Q4 when several subjects or a specific voice must stay consistent, or when the final has to be 4K. Use MiniMax H3 Max for quick start-frame clips and Gemini Omni 1.1 Flash when you need to edit or extend.

Oppgitt spørsmål

What is Vidu Q4?+

It is the newest video model from Vidu, released as a preview in October 2026. On ZOOOP it animates a start frame, or builds a scene from up to 12 reference images and up to 3 reference voices, with native audio.

Can I generate from text only?+

No. Vidu Q4 needs a start frame or at least one reference image. For pure text-to-video, use Gemini Omni 1.1 Flash or MiniMax H3 Max.

How do reference voices work?+

Upload up to three MP3 clips of 3 to 12 seconds each. The generated speech follows the voices in those clips, which keeps a character sounding the same across clips.

How long can a clip be?+

Any length from 3 to 16 seconds per generation.

Which resolutions and ratios are available?+

540p, 720p, 1080p, 2K and 4K. Reference mode offers 16:9, 9:16, 1:1, 4:3 and 3:4; start-frame mode follows the shape of your image.

Does it support an end frame?+

No, only a start frame. For start-and-end-frame transitions, use MiniMax H3 Max or Gemini Omni 1.1 Flash.