Access the Vidu Q4 Preview API on APIMart with model ID viduq4-preview. Vidu supports one first frame or 1–15 reference images with optional reference audio, producing 3–16 second clips at up to 4K. Generated audio and seed control are available; it does not support text-only generation or end-frame control.
Video link valid for 72 hours
Transparent pricing with no hidden fees. Pay only for what you use.
| Variant | Tier | GoldCurrent price | Platinum | Diamond | Official | Your savings |
|---|---|---|---|---|---|---|
viduq4-preview | 540p | 0.448Credits(~$0.0448)/sec | 0.4256Credits(~$0.0426)/sec | 0.4032Credits(~$0.0403)/sec | 0.56Credits(~$0.056)/sec | 20% |
| 720p | 0.912Credits(~$0.0912)/sec | 0.8664Credits(~$0.0866)/sec | 0.8208Credits(~$0.0821)/sec | 1.14Credits(~$0.114)/sec | 20% | |
| 1080p | 1.024Credits(~$0.1024)/sec | 0.9728Credits(~$0.0973)/sec | 0.9216Credits(~$0.0922)/sec | 1.28Credits(~$0.128)/sec | 20% | |
| 2K | 1.624Credits(~$0.1624)/sec | 1.5428Credits(~$0.1543)/sec | 1.4616Credits(~$0.1462)/sec | 2.03Credits(~$0.203)/sec | 20% | |
| 4K | 3.328Credits(~$0.3328)/sec | 3.1616Credits(~$0.3162)/sec | 2.9952Credits(~$0.2995)/sec | 4.16Credits(~$0.416)/sec | 20% |
* Actual costs are subject to final output.
Try Vidu Q4 in the playground or connect the Vidu API to your application. Start with one first frame and an optional prompt, or supply 1–15 reference images, a required prompt, and optional MP3 clips. Configure output from 540p to 4K and choose a duration of 3–16 seconds.
Explore Vidu Q4 inputs, output controls, and the asynchronous API workflow
Plan Vidu workflows around an approved first frame or reference materials
Practical answers about Vidu inputs, API parameters, pricing, and results
No. The Vidu Q4 Preview API needs one first frame for I2V or 1–15 images for R2V. Vidu R2V may also use up to 3 reference audio clips. Text-only generation and first-and-last-frame control are not supported.
For Vidu I2V, use first_frame_image, role first_frame, or one unmarked image without reference audio. For Vidu R2V, use role reference_image (alias reference), multiple images, or reference audio with unmarked images. Never mix an explicit first frame with references or reference audio. Image URLs and image_with_roles share a 15-image R2V limit. Each PNG, JPEG/JPG, or WEBP image must be at most 50 MB; a base64 request must be under 20 MB.
A prompt is optional for Vidu I2V and required for Vidu R2V, with a 20,000-character limit in either mode. Describe the action, scene, and how the supplied references should inform the video.
Vidu supports integer durations of 3–16 seconds (default 5) and 540p, 720p, 1080p, 2K, or 4K output (default 720p). For Vidu R2V, choose 1:1, 9:16, 16:9, 3:4, or 4:3; the default is 16:9. aspect_ratio takes precedence over its size alias. I2V ignores both fields and follows the first-frame image.
Vidu audio defaults to true; false produces silent output at the same price. R2V accepts 0–3 MP3 references, each 3–12 seconds and up to 50 MB. Upstream checks their format and duration; an invalid clip may cause a refunded task failure. Vidu seed accepts an integer; omit it or use 0 for random.
Both Vidu modes use duration multiplied by the selected resolution's current per-second price. Reference images and audio carry no material surcharge; generated audio on or off costs the same. Check live Vidu pricing for your account. Failed tasks are fully refunded.
The Vidu generation API at POST /v1/videos/generations returns a task_id. Poll GET /v1/tasks/{task_id} every 5–10 seconds until completed or failed. Find completed video URLs in result.videos[].url.
Save Vidu videos within 72 hours and check expiry information in the task response. A query timeout does not mean generation failed; resume querying the original task instead of submitting it again.
Explore more models in the same category.

Wan 3.0 Video Prime
Wan 3.0 Prime is the Prime version of Alibaba Cloud's all-in-one Wan 3.0 video model, with general, first-and-last-frame and file or webpage modes and 2–30 second output.

Gemini Omni 1.1 Flash
Gemini Omni 1.1 Flash is a multimodal video model officially released by Google that supports rapid generation, natural language editing, and cinematic-quality visual control. This page connects to the official service through APIMart’s unified video API and uses separate pricing and task logs.

MiniMax H3 Max
MiniMax H3 Max generates video from text and images at 480P, 768P or 1080P with 5–15 second clips, first/last frames, multimodal references and synced audio.

Grok Imagine Video
Grok Imagine Video 1.5 is xAI's official video model for text-to-video and image-to-video, animating still images from reference image URLs with high visual fidelity.
Guides, comparisons and pricing analysis from the APIMart blog.