ShengShu video model on Pixel Dojo
Vidu Q3 AI Video Generator
Vidu Q3 turns one prompt into a cinematic 1080p clip with its own soundtrack, and it can cut between two shots in a single generation.
Vidu Q3 is ShengShu's text to video and image to video model. On Pixel Dojo you can make clips from 1 to 16 seconds at up to 1080p, in five aspect ratios from widescreen 16:9 to vertical 9:16, with sound made in the same pass. We made every clip on this page with it, and each one comes with its exact prompt so you can start from ours.
10 credits ($0.63) per 5 s clip at 720p without audio. Included in every plan, from $10/month.
Already a member? Sign in
By Brian Lovett · Updated
What Vidu Q3 does best
Vidu Q3 comes from ShengShu, the team behind the Vidu AI app. Give it a written prompt, or a prompt plus one image to use as the first frame, and it returns a finished clip from 1 to 16 seconds long. Four resolutions (360p, 540p, 720p and 1080p) and five aspect ratios (16:9, 4:3, 1:1, 3:4 and 9:16) cover cinema, social feeds and vertical video from one model.
Its standout trick is editing. Write a prompt as Shot 1 and Shot 2 and Vidu Q3 cuts between them inside a single clip, like our lighthouse storm scene that moves from a lantern lit close up to a boat on the waves. You can storyboard a short beat in one prompt instead of stitching clips together.
Sound comes from the same pass as the picture. Switch audio on, end your prompt with the sounds you want, and the clip arrives with its own soundtrack. Audio does not change the price.
We made every clip on this page with Vidu Q3 on Pixel Dojo, and it shines on atmosphere, characters, fashion and stylized animation. Faces carry real emotion, silk and hair move with the wind, water sprays catch the light, and anime keeps a crisp cel shaded look.
What you can make with Vidu Q3
Slow motion action, a fashion reel, a two shot storm scene, anime and an emotional close up, all made on Pixel Dojo with audio on. Every tile shows the exact prompt and settings we used. Tap Use this prompt to start from any of them.
Action with real physics. Each hoofbeat throws up sunlit spray, the mane streams and the wet sand mirrors the sunrise across all 5 seconds. Start from this prompt for any slow motion hero shot.
Low tracking shot at water level. A black horse gallops straight through shallow surf at sunrise, each hoofbeat exploding sprays of seawater that catch the backlight like sparks, mane and tail streaming, wet sand mirroring a peach and gold sky. Long telephoto lens with heavy compression, slow motion feel, fine film grain. Sounds of thundering hooves, crashing waves and splashing water.
Fashion editorial in vertical 9:16 at 1080p. The crimson gown sweeps out in wide red arcs against the white salt flat, her hair whips in the wind and her face stays sharp from the first frame to the last. Made for vertical feeds.
Symmetrical full length shot, locked off camera. A tall woman dancer in a flowing crimson silk gown spins on a white salt flat under hard noon sun, the long fabric unfurling in a wide spiral and whipping in the wind, her black hair loose, her face turning toward the camera, one hard shadow beneath her. Minimal composition, bleached pastel sky, high fashion editorial look, 35mm film texture. Sounds of wind gusts, fabric snapping and a slow cello note.
Two shots, one generation. We wrote Shot 1 and Shot 2 in a single prompt and Vidu Q3 cut on its own, from the lantern lit close up to the fishing boat on black storm waves. Storyboard a short scene without stitching clips together.
Two shot scene with a hard cut. Shot 1: medium close up of a weathered lighthouse keeper in his sixties with a grey beard and a yellow oilskin coat, lit by a swinging storm lantern, he looks up at the window and says in a low gravelly voice, "She is coming in too fast." Shot 2: wide shot through the rain streaked lighthouse window of a small fishing boat pitching on huge black waves at night as the lighthouse beam sweeps across it. Cold blue storm palette with warm amber lantern light. Sounds of howling wind, rain on glass, creaking timber and a tense string score.
Stylized anime across 10 seconds. Crisp cel shading, hand painted rooftops and forks of lightning as the swordsman runs the temple ridge in the rain and leaps above a street of glowing lanterns.
2D anime film style, hand painted backgrounds and crisp cel shading. Low angle tracking shot. A grey bearded wandering swordsman in a straw hat and tattered indigo coat glances back over his shoulder, then sprints across wet temple rooftops in a thunderstorm and leaps a wide gap as lightning splits the sky behind him, coat snapping in the wind, rain falling in thick diagonal streaks, paper lanterns swaying in the streets below. Deep blue night palette with hot white lightning flashes. Sounds of pounding rain, rolling thunder, footsteps on clay tiles and a swelling taiko drum score.
Emotional close up at 540p for 5 credits. Amber and blue streaks of passing city light slide across her face as she lowers the letter, then she smiles with her eyes closed and a tear on her cheek.
Extreme close up, slow push in. An elderly woman with silver hair and deep laugh lines sits at a night train window, city lights sliding across her face in streaks of amber and blue, she lowers a folded letter and a single tear rolls down her cheek as she smiles. Soft practical light, 85mm lens, shallow depth of field, warm film grain. Sounds of train wheels on rails, a soft chime and a gentle piano melody.
Vidu Q3 settings for the look you want
These are the controls on the Vidu Q3 tool and the public API, with how we used each one.
- Resolution360p, 540p, 720p, 1080p (720p default)
- Use 1080p at 3 credits per second for full detail on a big screen. 540p costs 1 credit per second, the same as 360p, and is a low cost way to try a new idea or style.
- Duration1 to 16 seconds (the tool offers 3, 5, 8, 10 and 16; the API takes any whole number)
- Five seconds suits one action beat. Give a two shot prompt 8 seconds so each shot has room to land, and go to 10 seconds or longer for a run of action like our anime rooftop chase.
- Aspect ratio16:9, 4:3, 1:1, 3:4, 9:16
- Pick the frame before you generate. Our 9:16 fashion clip keeps her whole figure, head to feet, in the vertical frame.
- AudioOn or off (off by default), same credit price either way
- Turn it on and end the prompt with a Sounds of line naming three or four specific sounds. Leave it off if you will score the clip yourself. After you tap "Use this prompt", switch audio on in the tool so your clip gets its soundtrack too.
- ModeText to video, or image to video with your image as the first frame
- In image to video the source image sets the frame shape, so crop the still to the aspect you want first.
- Shot structureWritten in the prompt: Shot 1, Shot 2
- Give each shot one subject, one camera position and one action. That format gave us a clean cut from the lantern lit close up to the boat in the storm.
Vidu Q3 prompts: how to get results like these
The prompt habits behind every clip on this page, shown here on a 5 second anime rooftop run with forked lightning, driving rain and crisp cel shading, made at 540p for 5 credits.
Anime at 540p for 5 credits. He glances back under his straw hat, lightning forks over the temple roofs and he runs the ridge through driving rain. Start from this prompt for your own anime action scene.
2D anime film style, hand painted backgrounds and crisp cel shading. Low angle tracking shot as a grey bearded wandering swordsman in a straw hat and tattered indigo coat sprints across wet temple rooftops in a thunderstorm and leaps a gap as lightning splits the sky behind him, coat snapping in the wind, rain falling in thick diagonal streaks, paper lanterns swaying below. Deep blue night palette with hot white lightning flashes. Sounds of pounding rain, rolling thunder, footsteps on clay tiles and a swelling taiko drum score.
- Write the shot like a director. Open with the shot size and camera move, then one subject with specific details, one action, the light and the lens. Every prompt on this page follows that order.
- Cut inside one clip. Label the beats Shot 1 and Shot 2, give each one subject, one camera position and one action, and allow 8 seconds so both shots land, as our lighthouse storm scene does.
- Ask for sound by name. Switch audio on and end the prompt with a Sounds of line naming three or four specific sounds, like thundering hooves, crashing waves and wind.
- Go for motion and emotion. Expressive faces, flowing fabric, animals in motion and cel shaded anime are where Vidu Q3 shines, and every clip on this page shows one of them.
- Match the frame to the feed. Choose 9:16 for vertical video and 16:9 for widescreen before you generate, and crop a source image to the same shape in image to video mode.
Vidu Q3 pricing: credits per clip
Vidu Q3 is included in every Pixel Dojo plan, and plans start at $10/month. Each clip uses credits by length and resolution: 1 credit per second at 360p and 540p, 2 credits per second at 720p and 3 credits per second at 1080p. Audio does not change the price.
For the clips on this page that means 15 credits for a 5 second 1080p clip like the galloping horse or the fashion reel, 24 credits for the 8 second 1080p storm scene, 30 credits for the 10 second 1080p anime chase and 5 credits for the 5 second 540p close up. The tool default, 5 seconds at 720p, is 10 credits, and the longest clip, 16 seconds at 1080p, is 48 credits.
Vidu Q3 questions
What is Vidu Q3?
Vidu Q3 is ShengShu's AI video generator for text to video and image to video. It makes clips from 1 to 16 seconds at up to 1080p, with optional sound made in the same pass, and it can cut between shots inside one clip.
What is Vidu Q3 best at?
Cinematic character work, fashion, action with water and fabric, and stylized anime. On this page it lands a tearful smile in a tight close up, sweeps a silk gown through the wind, throws sunlit spray under a galloping horse and holds a cel shaded look across a 10 second rooftop chase.
How do I use Vidu Q3 on Pixel Dojo?
Open the Vidu Q3 tool, write or paste a prompt, pick the resolution, length, aspect ratio and audio, and generate. Tap Use this prompt on any clip here to start from our wording. Developers can send the same fields (prompt, mode, image_url, duration, resolution, aspect_ratio and generate_audio) to the Vidu Q3 model on the Pixel Dojo API, with mode set to image-to-video when you include an image_url.
How much does Vidu Q3 cost?
Vidu Q3 is included in every Pixel Dojo plan, from $10/month. A clip costs 1 credit per second at 540p, 2 credits per second at 720p and 3 credits per second at 1080p, so a 5 second 1080p clip is 15 credits. Audio comes at the same price.
How long can a Vidu Q3 video be?
From 1 to 16 seconds. The tool offers 3, 5, 8, 10 and 16 seconds, and the API accepts any whole number in that range. The clips on this page run 5, 8 and 10 seconds.
Does Vidu Q3 generate sound?
Yes. Switch audio on and Vidu Q3 builds a soundtrack from the sounds you name in your prompt, in the same pass as the picture and at the same credit price. Every clip on this page was made with audio on.
What resolution and aspect ratios does Vidu Q3 support?
360p, 540p, 720p and 1080p, in 16:9, 4:3, 1:1, 3:4 and 9:16. In image to video mode the shape of your source image sets the frame.
Make your first Vidu Q3 clip
Write a prompt above or start from any example on this page. One subscription covers every model on Pixel Dojo.
Start with Vidu Q3