> ## Documentation Index
> Fetch the complete documentation index at: https://docs.qingbo.dev/llms.txt
> Use this file to discover all available pages before exploring further.
# Vidu Q4 Preview
> ShengShu Vidu Q4 Preview: start-frame-to-video, or video from up to 15 reference images plus 3 reference audio clips, 3–16 seconds, up to 4K, audio on by default
ShengShu Technology's Vidu Q4 preview, model ID `vidu-q4-preview`. It has a single `generate` action with two ways in, chosen by the assets you send: a start frame for image-to-video, or reference images (optionally with reference audio) for reference-to-video, where characters and subjects come from the images. Output runs from `540p` to `4k`, with a dialogue and sound-effects track on by default. The model does not do text-only or first-and-last-frame video; for those, use `vidu-q3-pro` or `vidu-q3-turbo` from the [Vidu Q3 Series](/en/api-reference/video/vidu).
## Quick start
```bash Start frame to video theme={"system"}
curl -X POST https://api.qingbo.ai/v1/tasks \
-H "Authorization: Bearer $WAVE_API_KEY" \
-H "Content-Type: application/json" \
-H "Idempotency-Key: vidu-q4-demo-001" \
-d '{
"model": "vidu-q4-preview",
"action": "generate",
"prompt": "The person turns toward the window, slow push-in, soft indoor light",
"first_frame_image": "https://cdn.example.com/first-frame.jpg",
"duration": 5,
"resolution": "1080p"
}'
```
```bash Reference to video theme={"system"}
curl -X POST https://api.qingbo.ai/v1/tasks \
-H "Authorization: Bearer $WAVE_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "vidu-q4-preview",
"action": "generate",
"prompt": "The person from image 1 walks into the bookstore from image 2 and greets the clerk with the line from the reference audio",
"image_urls": [
"https://cdn.example.com/character.jpg",
"https://cdn.example.com/bookstore.jpg"
],
"audio_urls": ["https://cdn.example.com/line.mp3"],
"aspect_ratio": "9:16",
"duration": 8,
"resolution": "720p"
}'
```
```bash Single reference image theme={"system"}
curl -X POST https://api.qingbo.ai/v1/tasks \
-H "Authorization: Bearer $WAVE_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "vidu-q4-preview",
"action": "generate",
"prompt": "The person from the reference image runs down a street after the rain, camera tracking from the side",
"image_urls": ["https://cdn.example.com/character.jpg"],
"aspect_ratio": "16:9",
"duration": 5,
"resolution": "4k",
"generate_audio": false
}'
```
A successful submission returns a `task_id`. Retrieve the result with [`GET /v1/tasks/{task_id}`](/en/api-reference/task/status), wait inline with [`Prefer: wait`](/en/api-reference/task/submit#in-request-waiting-prefer-wait), or configure a webhook.
## Request parameters
Always `vidu-q4-preview`.
`generate` is the only action. The input mode follows the assets you send:
* Start frame to video: send `first_frame_image`; the prompt is optional
* Reference to video: send `image_urls`, optionally with `audio_urls`; the prompt is required
Sending neither returns `400`; the model does not do text-only video.
Description of the video — motion, camera, mood. Required when `image_urls` is sent; optional with `first_frame_image` alone, in which case the model decides the content from the start frame.
Start frame URL, used as the opening shot; the aspect ratio follows this image. With it set, `image_urls` and `audio_urls` cannot be sent.
Reference image URLs, 1–15; they decide characters, subjects, setting and style. Every image is treated as a reference, a single image included — it is never used as a start frame. With it set, `prompt` is required.
Reference audio URLs, at most 3 clips, MP3, 3–12 seconds each. Only usable together with `image_urls`, and cannot be sent with `first_frame_image`. If a clip's format or length does not fit, the task fails during generation and is refunded in full.
Clip length in seconds, an integer from `3` to `16`.
Output resolution: `540p`, `720p`, `1080p`, `2k` or `4k`.
Frame ratio: `16:9`, `9:16`, `4:3`, `3:4`, `1:1`. Applies to reference-to-video only; with `first_frame_image` it has no effect and the ratio follows the start frame.
Whether to generate an audio track (dialogue and sound effects). Set it to `false` for a silent clip; the price is the same.
Random seed. The same parameters with the same seed produce similar, though not identical, results.
The shared `callback_url`, `callback_events`, `Prefer: wait`, `Idempotency-Key` and cost-cap headers are documented in [Submit Task](/en/api-reference/task/submit).
The model does not accept `last_frame_image`, `video_urls`, `n`, `negative_prompt`, `watermark` or `quality`. Anything beyond the parameters above returns `400` and is not billed.
## Limits
| Condition | Limit |
| - | - |
| Per request | One video per task; up to 15 reference images and 3 reference audio clips. |
| Neither `first_frame_image` nor `image_urls` sent | Provide a start frame (first\_frame\_image) or 1-15 reference images (image\_urls); text-only generation is not supported. |
| `first_frame_image` sent | With a start frame, reference images and audio cannot be added, and the aspect ratio follows the image. |
| `image_urls` sent | Reference-to-video requires a prompt. |
| `audio_urls` sent | Reference audio needs at least one reference image. |
| Item | Start frame to video | Reference to video |
| - | - | - |
| Images | `first_frame_image`, 1 image | `image_urls`, 1–15 images |
| Reference audio | Not supported | `audio_urls`, at most 3 clips, MP3, 3–12 seconds each |
| Prompt | Optional | Required |
| Aspect ratio | Follows the start frame | `aspect_ratio`, default `16:9` |
| Duration and resolution | 3–16 seconds, `540p`–`4k` | 3–16 seconds, `540p`–`4k` |
## Pricing
Billed by **resolution tier times output seconds**: the resolution sets the per-second rate, `duration` sets the number of seconds, and their product is the cost of the call — known before you submit. A higher resolution has a higher per-second rate.
Start-frame and reference-to-video are priced the same. Reference images and reference audio are not billed separately; `generate_audio`, `seed` and the aspect ratio do not affect the price either, so audio and silent clips cost the same. Without `resolution` the `720p` default tier applies.
Rates are in `price_config` on `GET /v1/models`, and in the console's Model Market.
What a single call actually cost is the `cost` field on the task response — an integer quota at 500,000 quota = 1 USD.
You are only billed for a video that is produced. Failed and cancelled tasks, and tasks that return no usable video, are refunded in full. The final charge is the `cost` field on the task response — an integer quota, not dollars.
## Response
```json Task completed theme={"system"}
{
"task_id": "task-wave1791564845b950000001",
"model": "vidu-q4-preview",
"action": "generate",
"status": "completed",
"progress": "100%",
"created_at": 1791564845,
"completed_at": 1791564954,
"result": {
"videos": [
{
"url": ["https://cdn.example.com/result.mp4"],
"expires_at": 1792169754
}
]
},
"billing_status": "settled",
"cost": 75600,
"urls": {
"get": "https://api.qingbo.ai/v1/tasks/task-wave1791564845b950000001",
"cancel": "https://api.qingbo.ai/v1/tasks/task-wave1791564845b950000001/cancel"
}
}
```
`result.videos` holds the generated video; `url` is an array and `expires_at` is when the link stops working (Unix seconds) — copy the file to your own storage before then. Full field reference in [Query Task Status](/en/api-reference/task/status).
## Related
* [Vidu Q3 Series](/en/api-reference/video/vidu)
* [Video Generation Overview](/en/api-reference/video/overview)
* [Submit Task](/en/api-reference/task/submit)
* [Query Task Status](/en/api-reference/task/status)
* [Task System](/en/docs/task-system)
This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.