Vidu Q4 is ShengShu Technology's multimodal video model. It generates 3–16 second clips from a first-frame image or from image and voice references, with native dialogue and sound effects. On Molyin it outputs 540p, 720p, 1080p, 2K or 4K.

Which generation modes are available?

Image-to-video animates a single first-frame image and follows its aspect ratio. Reference-to-video combines image and voice references and lets you choose 16:9, 9:16, 4:3, 3:4 or 1:1.

How many reference assets can I use?

You can use up to 12 images and 3 voice clips in reference-to-video. Each voice clip must be an MP3 lasting 3–12 seconds. Reference-to-video does not take reference videos.

Can Vidu Q4 generate dialogue and stay silent when needed?

Yes. Native audio can include synchronized dialogue and sound effects, or you can turn audio off for a silent clip. Image-to-video supports both options, so a silent result is a real output choice, not a muted download.

How many credits does Vidu Q4 cost?

At 540p it costs 10 credits per second, so a 3-second clip is 30 credits. At 720p it costs 20/s, so 5 seconds is 100 and 16 seconds is 320. At 4K it costs 82/s, so a 3-second clip is 246. The full table above also includes 1080p (26/s) and 2K (40/s).

How should I write prompts for Vidu Q4?

Start with the subject, motion and camera behavior you want to see. For reference-to-video, give each asset an explicit role: @Image1 defines identity, @Image2 supplies a supporting subject or environment, and @Audio1 sets the voice. Keep one clear action per beat and state the aspect ratio in your settings rather than relying on the prompt to enforce it.

Should I use Vidu Q4 or Seedance 2.5?

Choose Vidu Q4 for first-frame animation and short scenes built from image and voice references, especially when you need 2K or 4K output. Choose Seedance 2.5 for longer 4–30 second sequences and larger multimodal reference stacks. Both use the same credit balance, so you can run the same concept on both and keep the stronger result.

Can I use Vidu Q4 videos commercially?

Yes. Videos generated on Molyin include standard commercial usage under our terms, including product ads, client work and monetized channels. You are responsible for the images and voice clips you upload, so use assets you have rights to.

Vidu Q4 — AI Video Generator

Vidu Q4

Animate a first frame or direct a scene with image and voice references. Vidu Q4 creates 3–16 second videos up to 4K with native dialogue and sound effects.

Failed generations auto-refunded
Credit packs never expire
Every model's price is public
Cancel anytime in one click

3–16 s
Clip length: Up to 4K
Five resolution tiers: 12 images + 3 voices
Reference assets: Dialogue + SFX
Native audio

Real Vidu Q4 outputs, complete prompts

Every clip below was generated with the same settings shown beside it. Choose Remix to load the full prompt and parameters into the generator, then adapt them for your own scene.

Vidu Q4
Reference to video
5s
720P

@Image1 rides @Image2 through a narrow rainy Tokyo alley at dusk, neon signs reflecting in the puddles. She rings the bell, smiles and calls out: "Sorry, coming through!" Tracking shot from the side; rain patters on her hood and tires splash through water.

Vidu Q4
Image to video
5s
720P

The paper crane flaps its folded wings, lifts off the notebook and glides toward the bright window as the hand pulls back in surprise. The camera tilts up to follow the crane through the morning light. Soft paper rustling, a quiet gasp and birdsong from outside.

Vidu Q4
Image to video
5s
720P

The cook lifts a ladle of broth from the steaming wok and pours it into a bowl, then looks at the customer and says with a grin: "Two more minutes, it is worth the wait." The customer laughs and nods. Slow push-in on the cook; oil sizzles, steam hisses and the crowd murmurs in the background.

Vidu Q4
Image to video
5s
720P

The husky turns its head toward the researcher and nuzzles their glove; the researcher scratches its neck and laughs softly. Snow keeps falling as the camera slowly orbits around the pair toward the lit cabins. Wind whistles across the snowfield and the husky gives a short happy bark.

Vidu Q4
Reference to video
5s
720P

On a sunny afternoon in a city park, @Image1 sits on a picnic blanket with @Image2 while @Image3 chases a butterfly across the grass. The robot hands her a cup of tea and she laughs. Handheld medium shot, soft breeze moving the leaves.

A cat and a robot walk into a chess game

Two reference images define the cat, the robot and their shared library setting. Vidu Q4 turns them into a coherent scene with warm lamplight, a small metallic voice and room-tone detail.

The prompt

In a cozy wood-paneled library, @Image1 sits across a chessboard from @Image2. The cat taps a knight forward with its paw; the robot's amber eyes blink, it tilts its dome head and says in a small metallic voice: "Well played." Warm lamplight, slow push-in, a clock ticks in the background.

The generated chess match — two reference subjects, one shared world, a clock ticking in the background.

What Is Vidu Q4?

Vidu Q4 is ShengShu Technology's multimodal video model. It animates a first-frame image and builds scenes from image and voice references, producing 3–16 second clips with synchronized dialogue and sound effects. Output ranges from 540p for quick drafts to 4K for detailed delivery.

Reference generation accepts up to 12 images and 3 voice clips. Each voice clip is an MP3 lasting 3–12 seconds, so you can guide identity, setting and delivery in one generation. Image-to-video follows the first frame's aspect ratio; reference-to-video offers 16:9, 9:16, 4:3, 3:4 and 1:1. Audio can be enabled or disabled to fit the edit.

On Molyin, Vidu Q4 sits inside the same AI video generator as our other models. There is one prompt box and one credit balance, with per-second rates by resolution shown below and on the pricing page.

Vidu Q4 vs Seedance 2.5

Both models handle images, references and native audio, but they optimize for different jobs. Here is how their current Molyin capabilities and credit rates compare.

Spec	Vidu Q4	Seedance 2.5
Clip length	3–16s	4–30s
Resolution	540p · 720p · 1080p · 2K · 4K	480p · 720p · 1080p
Reference assets	Up to 12 images + 3 voice clips	Up to 50 multimodal assets — 30 images, 10 videos, 10 audio
Audio	Native dialogue and sound effects, with silent output available	Voice, SFX and music generated with the picture
Credits per second	10 / 20 / 26 / 40 / 82 (540p / 720p / 1080p / 2K / 4K)	26 / 56 / 140 (480p / 720p / 1080p)
Best suited for	First-frame animation and short scenes built from image and voice references	Longer multi-shot sequences and large multimodal reference stacks

Both models share one credit balance on Molyin; rates shown are current execution rates.

Vidu Q4 vs models live on Molyin

Model	Positioning	Release Date	On Molyin
Vidu Q4	Image and reference video with native audio, up to 4K	Oct 7, 2026 (preview API)	Live — generate above
Kling 4.0	Long takes, keyframes and omni references	Oct 15, 2026 (general availability)	Coming soon
LTX 2.5	Open-weights video with native multi-shot and 4K	Aug 11, 2026 (open weights)	Live — generate above
Seedance 2.5	30-second takes directed second by second	Aug 7, 2026 (API)	Live — generate above
FLUX 3	Storyboard keyframes and take continuation	Aug 7, 2026 (video API)	Live — generate above
Wan 3.0	All-in-one multimodal generation, up to 30 seconds	Aug 6, 2026	Live — generate above
MiniMax H3	Omni-modal generation with native 2K and stereo	Jul 31, 2026 · open weights announced	Live — generate above

One subscription, one credit balance — every live model is included.

Vidu Q4 full specifications

Vidu Q4 full specifications
Parameter	Value	Notes
Generation modes	Image-to-video and reference-to-video	Animate a first frame or compose a scene from image and voice references.
Resolution tiers	540P / 720P / 1080P / 2K / 4K	Five output tiers from quick drafts to detailed masters.
Clip length	3–16 s	Choose any whole-second duration in the supported range.
Aspect ratios	16:9 · 9:16 · 4:3 · 3:4 · 1:1	Image-to-video follows the first frame; reference-to-video offers all five ratios.
Reference assets	Up to 12 images + 3 voice clips	Each MP3 voice clip lasts 3–12 seconds; reference video input is not used.
Audio	Native dialogue and sound effects	Keep synchronized sound or choose silent output for post-production.
Credits per second	10 / 20 / 26 / 40 / 82	Rates correspond to 540p / 720p / 1080p / 2K / 4K output.

Everything listed here is already supported on the site, and stays in sync with the model's latest official capabilities.

Vidu Q4 credit pricing

Pay per second by resolution. Use the same credit balance for Vidu Q4 and every other model — failed generations are refunded automatically.

AI video model credit rates by resolution — credits per second and per-clip examples
Model	Resolution	Credits/sec	5s Video	5s Video + 3s Reference
Vidu Q4	540p	10	50 credits	Reference images only
	720p	20	100 credits	Reference images only
	1080p	26	130 credits	Reference images only
	2k	40	200 credits	Reference images only
	4k	82	410 credits	Reference images only

Subscription plans and one-time credit packs are compared side by side on the pricing page.

Compare credit rates for all models →

How to generate with Vidu Q4

Three steps from reference to finished clip

Choose a mode

Use image-to-video to animate a first frame, or reference-to-video to compose from images and voice clips.

Direct the scene

Write the motion and setting, then use @Image1 and @Audio1 to give each reference a clear role.

Generate and export

Select resolution, duration and aspect ratio, preview the result, then download the original file.

Vidu Q4 FAQ

Common questions about inputs, audio, pricing and choosing between Vidu Q4 and Seedance 2.5.

Start Creating with Molyin Today

Start Creating Free View Pricing