Vidu Q3 Pro Fast · image-to-video only

Vidu Q3 Pro Fast Image to Video Generator

One uploaded frame in, one clip out. This route fixes the tier and the input, then shows the official per-second rate next to the point estimate the workspace calculates for it.

Try Vidu Q3 Pro Fast
01

Run Vidu Q3 Pro Fast in the workspace

Vidu Q3 Pro Fast loads with image-to-video as its only input route. Upload one frame, set a whole number of seconds between 2 and 15, choose 720p or 1080p, and read the point estimate before you submit.

LIVE
5s
2s15s

This model does not support reference video input

Results

Creation history

Generated content will appear here
02 / Overview

Vidu Q3 Pro Fast rates and the single route it runs

Facts reviewed:

Provider
Vidu
Official model ID
viduq3-pro-fast
Published route
Image to video only
Documented on
Official pricing table and image-to-video API reference
Not listed on
The official Q3 model map page
Official resolution
720p default, 1080p optional; no 540p tier
Official duration
1 to 16 seconds, default 5
Route on this page
Image to video
Duration here
2 to 15 whole seconds
Official rate at 1080P
25 credits per second, $0.125
Official rate at 720P
20 credits per second, $0.10

One published route against the switches this page does not expose

Vidu documents Pro Fast in two places, and both describe a model with a single input route and no 540p tier. This workspace adds its own narrower duration range on top of that.

OFFICIAL

Documented on the official Vidu pages

  • viduq3-pro-fast is typed as img2video on the official pricing table, with no text-to-video, start-end2video or reference2video row under that model ID
  • The image-to-video reference gives its resolution options as 720p by default with 1080p available, in a bullet list where Pro and Turbo both also list 540p
  • A duration of 1 to 16 seconds with a default of 5, quoted for Pro Fast, Pro and Turbo together in one row of the same parameter table
  • Described as efficiently generating high-quality audio and video content with faster generation speed and high cost-effectiveness
  • Direct audio-video output defaults to on for the Q3 series, and off-peak scheduling runs at half rate with a 48 hour generation window
LIVE PLATFORM

What this workspace exposes for the same model

  • A single input route: image to video. Text, first and last frame, and reference-to-video are not selectable on this page
  • One uploaded frame, used as the starting frame of the clip. The reference documents a single image for this parameter
  • A resolution pool of 720p and 1080p, defaulting to 720p, which matches the two values the official reference lists
  • Duration from 2 to 15 whole seconds, narrower than the documented 1 to 16 range at both ends
  • No audio switch. Vidu defaults Q3 models to generated audio, and this route exposes no control for it
  • No off-peak switch. The half-rate scheduling Vidu documents for Q3 is not selectable from this page

Two official pages describe this model and neither describes this page. The pricing table gives one typed route and two resolution rows; the image-to-video reference gives the parameter behaviour. The model map page does not list viduq3-pro-fast at all, so nothing on this page leans on its capability columns. There is also a difference of level between Vidu documents worth naming: the function list describes image-to-video as supporting background music and motion amplitude, while the parameter notes on the same site state that background music is not available on Q3 models and that motion amplitude has no effect on Q2 and Q3. Read together, the function list describes the function and the parameter notes describe the model. This page follows the parameter notes and says so, rather than repeating the broader function-level promise as if it applied here.

Model brief

What the Fast label covers and what it leaves alone

Vidu Q3 Pro Fast is the narrowest of the three Q3 tiers this platform opens. Where Pro and Turbo are typed with three input routes on the official pricing table, Pro Fast is typed with image-to-video alone, and the same table gives it no 540P row while both siblings have one. For its own part the official model map page does not list the model ID at all. So the picture from the documents is consistent rather than flattering: this is a route built around a single kind of input, and it is priced above the tier that does more.

The name deserves a careful reading. Vidu describes the model as efficiently generating high-quality audio and video content with faster generation speed and high cost-effectiveness, and that wording is where most of the search interest in this model comes from. Two things are worth stating plainly about it. First, the wording names no baseline, so it does not establish that Pro Fast is faster than Pro, and Vidu publishes no latency or speed multiplier anywhere in the documents this page relies on. Second, the same official table places Pro Fast above Pro at 1080p, at 25 credits per second against 24, and far above Turbo at 13. The official description and the official price list therefore point in different directions, and both are official. This page lists them side by side and does not pretend they resolve into a single tidy claim.

What the page does instead is make the one available route work properly. It documents the parameter behaviour that applies to a single starting frame, states the missing 540p tier as a fact confirmed by two independent official pages, shows the arithmetic that connects this tier to its two siblings on the same table, and then names every control the workspace does not expose so that nothing here reads as an implied capability. Where Vidu leaves something undefined, the page leaves it undefined too.

03 / Capabilities

Setting up an image-to-video task on this route

Everything below maps to a parameter the official image-to-video reference documents for viduq3-pro-fast, or to a control this workspace actually renders.

01

One starting frame as the whole visual input

The reference accepts a single image for the images parameter and describes it as the start frame of the generated clip. Accepted formats are png, jpeg, jpg and webp, each image is limited to 50MB, the request body must stay under 20MB, and the aspect ratio of the supplied image has to be within 1:4 and 4:1 rather than an extreme panorama or column.

02

A prompt of up to 5000 characters

Prompts are optional in the reference and capped at 5000 characters, which is generous enough for shot description, camera behaviour, lighting and the timing of the action inside a single clip. Because the starting frame already fixes composition, the prompt is where motion, pacing and atmosphere get defined.

03

Five aspect ratios

Choose 16:9, 9:16, 3:4, 4:3 or 1:1. The two portrait and landscape ratios outside the usual pair are documented for Q2 and Q3 models specifically, so their availability here is a documented Q3 capability rather than a general platform default.

04

720p or 1080p output

Two resolutions, with 720p as the default. These are exactly the two values the official reference lists for this model, and they are also why this route has no low-resolution warm-up setting: 540p exists on the official table for the other two Q3 tiers, and the workspace does not offer it here because the model does not document it.

05

Whole-second duration from 2 to 15

Set any whole number of seconds in that range, with 5 as the documented API default. The workspace range is narrower than the official 1 to 16 at both ends, so a one-second test clip and a full sixteen-second take are both outside what this page can produce.

06

A point estimate before submitting

The workspace shows what the task will cost on this platform before you commit to it. The number follows the platform schedule rather than the official Vidu rate, so the two are worth reading as separate figures: the official rows tell you how the tiers compare, the estimate tells you what this submission costs here.

07

Reproducible runs through the seed value

The reference documents a seed parameter that defaults to random and can be set manually, with a manually set value overriding the random default. Combined with a fixed starting frame and a fixed prompt, that is the closest thing to a repeatable run this route offers, though the reference does not promise identical output for identical seeds.

04 / Scene ideas

Jobs where a single still frame is the whole input

Everything below maps to a parameter the official image-to-video reference documents for viduq3-pro-fast, or to a control this workspace actually renders.

Product stills that need a camera move

A catalogue or studio photograph already carries the shape, label and colour that have to survive into the clip. Generating from that frame keeps the product consistent and spends the whole prompt on the movement instead of on describing the object.

Book covers, posters and key art

Cover art arrives as a finished composition with typography already placed. Starting from it produces an opening beat for a trailer or a social post without re-typesetting the artwork or risking a garbled recreation of the lettering.

Illustration and character frames

A single drawn or painted panel sets the character design, the palette and the line quality. Animating that panel keeps the design under your control and lets the prompt concentrate on how the figure should breathe, turn or shift weight.

Landscape and travel photographs

A wide photograph with clear depth cues is a good candidate for slow parallax, drifting cloud, moving water or a gradual light change. The composition is already decided, so the clip only has to add motion that respects it.

Architectural and interior frames

Renders and interior photography benefit from restrained, continuous movement. Because the starting frame fixes perspective, the prompt can specify a single deliberate camera path rather than asking the model to invent the space.

Storyboard frames that need to move

When a storyboard is already approved, animating its frames keeps the edit you signed off on. Each clip becomes one beat in the sequence, and the frame itself is the contract about what the viewer sees.

05 / Prompt lab

Prompt starters that begin from an uploaded frame

Every starter below assumes image-to-video with a single frame already uploaded, and none of them depends on a control this route does not expose.

SCENE 01image to video

Slow orbit around a studio product still

Tests how faithfully a product frame survives a camera move, which is the main risk when animating catalogue photography.

View prompt
Starting from the uploaded image, move the camera slowly to the right in a shallow arc of about 25 degrees, keeping the product centred and the label facing the lens throughout. Keep studio lighting and the backdrop pattern fixed. Do not change the product shape, colour or proportions, and do not add reflections or hands.
SCENE 02image to video

Cloud drift over a wide landscape frame

A low-risk first test: the shot only needs the sky to move while the ground stays still and sharp.

View prompt
From the uploaded photograph, drift the high clouds slowly from left to right while the land, water and foreground remain completely still. Late afternoon light stays constant. Movement should be smooth and continuous with no camera shake and no change to the horizon line.
SCENE 03image to video

Character panel taking a breath

A restrained motion test for illustrated frames, where the design has to survive untouched while one small movement reads clearly.

View prompt
Using the uploaded illustration as the opening frame, have the character take one slow breath: the chest rises slightly, the shoulders settle, and the head tilts a few degrees. Keep the line quality, palette and background exactly as drawn. No camera movement, no new elements, no text.
SCENE 04image to video

Interior render with a walking camera

Tests whether a single still can carry a deliberate camera path through a space without the geometry warping.

View prompt
Starting from the uploaded interior render, push the camera forward slowly along the centre line toward the far window, at a steady walking pace. Keep verticals straight and keep the furniture geometry unchanged. Daylight from the window stays constant, and no people enter the frame.
SCENE 05image to video

Poster art with a subtle light sweep

Keeps placed typography intact while adding the smallest possible motion, which is usually what a poster animation needs.

View prompt
From the uploaded poster, sweep a soft highlight slowly across the artwork from top left to bottom right, as if a light passed over a printed surface. All typography, edges and colours stay exactly as they are. No parallax, no scaling, no texture change.
SCENE 06image to video

Water surface moving in a still landscape

A motion-only brief where the frame is already finished and only one element in it should move.

View prompt
Using the uploaded landscape as the start frame, animate only the surface of the water with a gentle continuous ripple. Everything else in the frame stays frozen: mountains, trees, buildings and the reflection pattern stay in place. Camera does not move. Keep the colour and exposure unchanged.
COMPARE

Vidu Q3 Pro, Pro Fast and Turbo priced on one table

Every row is read from the official Vidu pricing table and the official model map, so the three columns are directly comparable. The gap rows are arithmetic on those published numbers, not additional official claims.

Vidu Q3 Pro, Pro Fast and Turbo priced on one table
ItemVidu Q3 ProVidu Q3 Pro FastVidu Q3 Turbo
Published routeimg2video, text2video, start-end2videoimg2videoimg2video, text2video, start-end2video
Official resolution options540p, 720p, 1080p720p, 1080p540p, 720p, 1080p
Official duration1 to 16 seconds1 to 16 seconds1 to 16 seconds
Listed on the model mapYesNoYes
Rate at 1080P24 credits per second, $0.1225 credits per second, $0.12513 credits per second, $0.065
Rate at 720P20 credits per second, $0.1020 credits per second, $0.1011 credits per second, $0.055
Rate at 540P9 credits per second, $0.045Not published7 credits per second, $0.035
Off-peak rate at 1080P12 credits per second, $0.0613 credits per second, $0.0657 credits per second, $0.035
Off-peak rate at 720P10 credits per second, $0.0510 credits per second, $0.056 credits per second, $0.03
Gap against Pro at 1080PDefines the baselineAbout 4 percent above ProAbout 46 percent below Pro
Gap against Turbo at 1080PAbout 85 percent above TurboAbout 92 percent above TurboDefines the baseline
Resolution pool on this platform720p, 1080p720p, 1080p720p, 1080p
LIVE PRICING

Official credits per second on the Pro Fast route

Vidu bills its own API per second and prices a credit at $0.005 on the official page. The point estimate in the workspace follows this platform schedule instead, so read the official rows as the tier comparison and the estimate as what a task costs here.

View platform pricing
01

How the estimate moves

Resolution and duration both move the estimate, and resolution moves it hardest. At 720p the official Pro Fast rate matches Pro exactly at 20 credits per second, so the premium over Pro appears only at 1080p, where Pro Fast is 25 against Pro at 24. Switching the resolution before submitting is the single change that most affects what this task costs.

02

Why 540p is not an option here

The official image-to-video reference lists resolution options per model in one bullet list, and the Pro Fast entry stops at 720p and 1080p while the Pro and Turbo entries both include 540p. The pricing table agrees, carrying no 540P row for Pro Fast. Two independent official pages leave out the same tier, so the omission is a documented property of the model rather than a gap in this workspace.

06 / Workflow

From upload to point estimate in five steps

  1. 01

    Confirm the route is image to video

    This page opens one route only. If the task starts from a written brief rather than an existing frame, or if it needs a specific closing image, the input type does not match what this tier publishes and the decision is better made before uploading anything.

  2. 02

    Upload the frame you want to start from

    One image, in png, jpeg, jpg or webp, within the official size ceiling and with an aspect ratio inside the documented 1:4 to 4:1 band. The frame is the composition, so choose it as you would choose a finished shot rather than a rough reference.

  3. 03

    Write only the motion into the prompt

    Since composition is already fixed by the frame, describe camera path, subject movement, pacing, lighting change and what must stay untouched. Restating the object is wasted effort and increases the chance of the model altering it.

  4. 04

    Set duration and resolution

    A whole number of seconds from 2 to 15, then 720p or 1080p. These two settings carry almost all of the cost difference on this route, and at the lower resolution the tier itself stops being a premium at all.

  5. 05

    Read the estimate, then submit

    Check the point estimate against the official per-second rows, and expect queueing behaviour rather than instant starts: Vidu documents a concurrency ceiling per organisation with excess tasks queued roughly in submission order.

Review before publishing

Where the Fast route stops and why

There is no 540p setting on this route, and that is a property of the model rather than a limitation of the workspace. The official image-to-video reference lists resolution options per model inside a single bullet list, and the entry for viduq3-pro-fast reads 720p with 1080p optional, while the entries for Pro and Turbo read 540p, 720p and 1080p. The official pricing table confirms the same absence by carrying no 540P row for this model ID. Two independently maintained official pages omit it, which is why the option is not offered here.

Duration is narrower here than the official range at both ends. Vidu documents 1 to 16 seconds with a default of 5, and this workspace exposes whole seconds from 2 to 15. That means a one-second test render is not available as a cheap sanity check, and neither is a full sixteen-second take. Plan the shot for the range that actually exists rather than for the documented one.

Two controls that Vidu documents for Q3 are absent from this page. Direct audio-video output defaults to on for the Q3 series, and this route exposes no switch for it, so the audio behaviour documented for the model is not a parameter you can set here. Off-peak scheduling is documented as a half-rate mode for Q3 with a 48 hour generation window and automatic cancellation with a refund for tasks that do not finish, and this route exposes no off-peak control either. Both are official capabilities that this page cannot offer, and both are stated rather than glossed over.

Finally, the Fast label should not be read as a published performance claim. The official description says faster generation speed and high cost-effectiveness without naming a baseline, and no latency figure, speed multiplier or throughput number appears in the documents this page relies on. The official FAQ is not a substitute for those numbers either: its duration and resolution answers still describe 360p, 720p and 1080p with four and eight second output, which contradicts the Q3 tables, so this page does not use it for specifications. Vidu publishes no discount or promotion for the Q3 range, and no third-party price, credit pack or ranking is used anywhere on this page.

Sources

Vidu documents behind this page

Capabilities, routes and rates change, and Vidu maintains a separate page for each of them. Every official page listed here was reachable without a login when this page was last reviewed, and the FAQ entry is listed because this page records a conflict in it rather than relying on it.

07 / FAQ

Vidu Q3 Pro Fast questions, answered

Does Vidu Q3 Pro Fast support text to video?
No. The official pricing table types the model ID as img2video only, and this route exposes image to video as its single input. A written brief with no starting frame belongs on a different tier, such as Vidu Q3 Pro or Vidu Q3 Turbo.
What resolution options are available here?
720p by default and 1080p optionally, which are exactly the two values the official image-to-video reference lists for this model. The 540p tier that the same reference documents for Pro and Turbo is not offered here.
How long can a clip be on this route?
Between 2 and 15 whole seconds. Vidu documents 1 to 16 seconds for the model with a default of 5, so this page is narrower at both ends and does not claim the official range as a local capability.
What does 25 credits per second cost in dollars?
At the official credit price of $0.005, the 1080P rate of 25 credits per second works out to $0.125 per second, which is the figure the official pricing table lists alongside it.
Is Pro Fast more expensive than Vidu Q3 Pro?
At 1080p it is, at 25 credits per second against 24. At 720p the two are identical at 20 credits per second, so the premium over Pro exists at the higher resolution only and is roughly four percent there.
How much more does it cost than Vidu Q3 Turbo?
At 1080P the official rates are 25 credits per second against 13, which is about 92 percent more. Turbo also has a 540P row at 7 credits per second that Pro Fast does not publish at all.
Is Pro Fast actually faster than Vidu Q3 Pro?
Vidu describes it as having faster generation speed and high cost-effectiveness, but the wording names no baseline and no latency or speed multiplier is published in the documents this page uses. This page does not assert a speed comparison that the official documentation does not state.
Can I supply first and last frames?
No. First and last frame generation is part of the start-end2video route, which the official pricing table lists for Pro and Turbo but not for this model ID, and this route exposes no control for it.
Can I add reference images for a character or product?
Not on this page. Reference-to-video is a separate published route, and the workspace does not open it here, so a task that depends on multiple reference images needs a different model page.
Does the model support background music?
Vidu states that background music is not available on Q3 models, and this route exposes no control for it. The function list describes background music at the function level, but the parameter notes restrict it per model, and the model level is what applies here.
Is off-peak scheduling available on this route?
The official table publishes off-peak rates for this model, including 13 credits per second at 1080P, but this page exposes no off-peak control. Off-peak is documented for Q3 with a 48 hour window and automatic refunds for unfinished tasks.
Is this model listed on the official Vidu model map?
No. The model map page does not list viduq3-pro-fast, while the official pricing table and the image-to-video API reference both do. This page relies on those two documents and does not attribute model map capability columns to it.

Other routes that take a different input

01

Vidu Q3 Pro

The sibling tier with three input routes, one credit per second more at 1080p and identical pricing at 720p.

02

Vidu Q3 Turbo

The lowest official per-second rate of the three tiers, with three input routes and a published 540P option.

03

Vidu Q3

The tier selector entry point, which runs Pro, Turbo and Pro Fast from one workspace with audio and off-peak controls.

04

Seedance 2.0 Fast

A speed-oriented route from another provider, useful as a cross-provider reference when comparing fast tiers.

Render your next scene

Start from a still frame

Return to the workspace, upload the frame you want to animate, confirm the duration and resolution, and check the estimate against the official rate before you generate.

Open Vidu Q3 Pro Fast workspace