Vidu’s lightweight video generation model, which quickly generates videos.
2026-02-03
Video Generation
Pricing:
$0.005 /Pointstarting from
Bulk order? Contact your manager for exclusive deals
API Overview
Vidu Q2 Pro Fast is a lightweight video-generation model product launched by Vidu (Shengshu Technology). Its core positioning is a “high-speed, high-quality short-video generation engine driven by single images or dual frames,” specifically designed to provide an efficient solution for rapid iteration, marketing content, and dynamic visual storytelling.
Ultra-fast generation: It boasts significantly faster production speeds than the standard Q2 Pro model, striking a balance between speed and image quality.
Two-mode support: It includes two sub-models: Image-to-Video Fast (single-image-driven) and Start-End-to-Video Fast (transition between start and end frames), covering diverse needs from static activation to smooth scene transitions.
Object-aware animation: It preserves facial features, hand movements, and fine-texture details, avoiding common distortions or artifacts and ensuring consistent subject representation.
Cinematic camera movements: It comes with built-in camera-path estimation, automatically simulating camera motions such as zooming, panning, and tracking shots to enhance the dynamism of the footage.
Out-of-the-box experience: It supports automatic background music addition, motion-amplitude adjustment (small/medium/large), and customizable durations from 1 to 8 seconds, perfectly matching the pacing of social-media platforms.
Applicable scenarios: It can be used for rapid content creation in industries such as e-commerce, advertising, and gaming, dramatically shortening video-production cycles.
🎬 Single-image vividization (Image-to-Video Fast):Upload a single image and describe the desired action (e.g., “a model confidently strutting down the runway”), and the model will generate a high-quality atmospheric video with natural limb movements and realistic fabric physics.
Widely used in fashion showcases, product animations, storyboarding previews, and other scenarios, enabling “static images to instantly transform into short videos.”
🔄 Seamless transition (Start-End-to-Video Fast):Provide a starting frame and an ending frame, and the model will intelligently interpolate to generate smooth transitions (e.g., clothing changes or morphing shapes), while keeping the background stable.
Ideal for creative expressions such as before-and-after comparisons, concept evolution, and magical effects.
🎛️ Fine-tuning options:The movement_amplitude parameter allows you to adjust the intensity of movements, while seed ensures reproducible results, and resolution supports switching between 720p and 1080p.
A built-in Prompt Enhancer automatically optimizes motion descriptions, lowering the barrier to creating effective prompts.
A new off_peak parameter is added, which is currently only supported by viduq3-pro.
Note: When enabling the off-peak mode, the audio must be set to true at the same time for normal billing.
Type
Model Version
Duration
Clarity
Style
Points Consumption
Off-Peak Points Consumption
Image-to-Video
viduq3-pro
1-16 seconds
1080p
General
32 points per second
16 points per second
Image-to-Video
viduq3-pro
1-16 seconds
720p
General
30 points per second
15 points per second
Image-to-Video
viduq3-pro
1-16 seconds
540p
General
14 points per second
7 points per second
Type
Model Name
Resolution
Points
Image-to-Video
viduq2-pro-fast
720p
8 points for the 1st second, +2 points for each subsequent second
Image-to-Video
viduq2-pro-fast
1080p
16 points for the 1st second, +4 points for each subsequent second
🖼️ Images
Supports inputting two images. The first uploaded image is regarded as the first frame, and the second image is regarded as the last frame. The model will generate a video with the images passed in this parameter.
Note:
The resolutions of the two input images (first and last frames) should be similar. The ratio of the resolution of the first frame image to the last frame image should be between 0.8 and 1.25. In addition, the image ratio needs to be less than 1:4 or greater than 4:1.
Supports passing image Base64 encoding or image URL (ensure accessibility).
Image supports png, jpeg, jpg, webp formats.
The image size does not exceed 50 MB.
promptstringOptional
✍️ Text Prompt
Text description for video generation.
Note:
The character length cannot exceed 2000 characters.
durationintegerOptional
⏱️ Video Duration Parameter (seconds)
The default value depends on the model:
🎵 Whether to add background music to the generated video
Default Value: false
Description:
When passing true, the system will automatically select suitable music from the preset BGM library and add it; if not passed or false, no BGM will be added.
BGM has no time limit, and the system will automatically adapt according to the video duration.
The BGM parameter does not take effect when the duration of the q2 model is 9 or 10 seconds.
