> ## 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 Q3 Series

> ShengShu Vidu Q3 — reference-to-video / text-to-video / image-to-video / first-last frame / multi-image remix, smart shot transitions

ShengShu Vidu Q3 video generation series. **Reference-to-video (R2V)** is the flagship capability — accepts 1-7 public images as subject / style references, with the prompt focused on motion and camera while the appearance comes from the reference images. Best fit for **character consistency / style continuity / multi-shot transitions**.

Four tiers:

* **`viduq3-pro`** — flagship tier, **auto-routes T2V / I2V / first-last frame by `image_urls` count**, audio output enabled by default (dialog + SFX)
* **`viduq3-mix`** — overall quality king, supports **multi-clip video / image remix** for coherent video output, joint audio generation, 720P / 1080P, ideal for short-drama re-cutting and ad creative remix
* **`viduq3`** — standard tier, the default daily choice, smarter shot transitions, 540p / 720p / 1080p, 3-16 sec
* **`viduq3-turbo`** — lightweight fast tier, fastest generation and lowest unit price, slightly lower quality than standard, ideal for previews / batch / A/B testing

Billed by **resolution × duration**, **per-second pricing**, any integer 1-16 sec.

## Pricing

| Model          | 540P            | 720P             | 1080P            |
| -------------- | --------------- | ---------------- | ---------------- |
| `viduq3-pro`   | `$0.0595` / sec | `$0.1275` / sec  | `$0.136` / sec   |
| `viduq3-mix`   | —               | `$0.10625` / sec | `$0.1275` / sec  |
| `viduq3`       | `$0.0425` / sec | `$0.085` / sec   | `$0.10625` / sec |
| `viduq3-turbo` | `$0.034` / sec  | `$0.051` / sec   | `$0.0595` / sec  |

## Examples

<CodeGroup>
  ```bash Text-to-video (Pro) theme={"system"}
  curl -X POST https://www.qingbo.dev/v1/tasks \
    -H "Authorization: Bearer $WAVE_API_KEY" \
    -H "Content-Type: application/json" \
    -d '{
      "model": "viduq3-pro",
      "prompt": "Sunset on the beach, a girl runs toward the setting sun with hair flowing, cinematic lighting",
      "duration": 5,
      "resolution": "1080p",
      "aspect_ratio": "16:9"
    }'
  ```

  ```bash Image-to-video (Pro single image) theme={"system"}
  curl -X POST https://www.qingbo.dev/v1/tasks \
    -H "Authorization: Bearer $WAVE_API_KEY" \
    -H "Content-Type: application/json" \
    -d '{
      "model": "viduq3-pro",
      "prompt": "The character slowly turns around, camera zooms in",
      "image_urls": ["https://cdn.example.com/portrait.jpg"],
      "duration": 5,
      "resolution": "1080p"
    }'
  ```

  ```bash First/last frame (Pro two images) theme={"system"}
  curl -X POST https://www.qingbo.dev/v1/tasks \
    -H "Authorization: Bearer $WAVE_API_KEY" \
    -H "Content-Type: application/json" \
    -d '{
      "model": "viduq3-pro",
      "action": "first_last_frame",
      "prompt": "Smooth transition from day to night with natural lighting changes",
      "first_frame_image": "https://cdn.example.com/day.jpg",
      "last_frame_image": "https://cdn.example.com/night.jpg",
      "duration": 8,
      "resolution": "1080p"
    }'
  ```

  ```bash Multi-image reference (standard R2V) theme={"system"}
  curl -X POST https://www.qingbo.dev/v1/tasks \
    -H "Authorization: Bearer $WAVE_API_KEY" \
    -H "Content-Type: application/json" \
    -d '{
      "model": "viduq3",
      "action": "reference",
      "prompt": "Keep the character look, run through the forest, camera follows",
      "image_urls": [
        "https://cdn.example.com/char-1.jpg",
        "https://cdn.example.com/char-2.jpg",
        "https://cdn.example.com/style-ref.jpg"
      ],
      "duration": 6,
      "resolution": "1080p",
      "aspect_ratio": "16:9"
    }'
  ```
</CodeGroup>

<ResponseExample>
  ```json Submission accepted theme={"system"}
  {
    "task_id": "task-wave1775285160b950328499",
    "model": "viduq3-pro",
    "action": "generate",
    "status": "queued",
    "created_at": 1775285160040,
    "progress": 0
  }
  ```
</ResponseExample>

After submission, poll status with [`GET /v1/tasks/{task_id}`](/en/api-reference/task/status); see [Task System](/en/docs/task-system) for details.

## Available Models

| Model ID       | Resolution          | Duration | Supported actions                           | Notes                                                   |
| -------------- | ------------------- | -------- | ------------------------------------------- | ------------------------------------------------------- |
| `viduq3-pro`   | 540p / 720p / 1080p | 1-16 sec | generate / image2video / first\_last\_frame | Flagship tier, audio output                             |
| `viduq3-mix`   | 720P / 1080P        | 1-16 sec | reference / image2video                     | Overall quality king, multi-asset remix                 |
| `viduq3`       | 540p / 720p / 1080p | 1-16 sec | reference / image2video                     | Standard tier, default daily choice                     |
| `viduq3-turbo` | 540p / 720p / 1080p | 1-16 sec | reference / image2video                     | Lightweight fast tier, fastest generation, lowest price |

## Common Parameters

<ParamField body="model" type="string" required>
  Model ID; see the [Available Models](#available-models) table
</ParamField>

<ParamField body="action" type="string">
  Operation type. Pro tier **may be omitted** (auto-routed by `image_urls` count); other tiers should specify it explicitly. Allowed values:

  * `generate` — text-to-video (Pro only)
  * `image2video` — image-to-video
  * `reference` — reference-to-video (default for standard / Mix / Turbo)
  * `first_last_frame` — first/last frame (Pro only)
</ParamField>

<ParamField body="prompt" type="string" required>
  Video description text. Focus on **motion / camera / atmosphere** — character appearance is determined by `image_urls`
</ParamField>

<ParamField body="aspect_ratio" type="string" default="16:9">
  Frame aspect ratio. Allowed values:

  * `16:9` — landscape widescreen
  * `9:16` — portrait
  * `4:3` — landscape
  * `3:4` — portrait
  * `1:1` — square
</ParamField>

<ParamField body="resolution" type="string" default="720p">
  Output resolution; see [Available Models](#available-models) for per-model support
</ParamField>

<ParamField body="duration" type="integer" default="5">
  Video duration in seconds, any integer **1-16**. Mix tier supports 1-second short videos
</ParamField>

<ParamField body="image_urls" type="string[]">
  Reference image URL array, **1-7 images**:

  * Standard / Mix / Turbo: subject / style references (R2V)
  * Pro tier: 0 = T2V / 1 = I2V / 2 = first-last frame; see [Auto-routing](#auto-routing)
</ParamField>

<ParamField body="first_frame_image" type="string">
  First-frame image URL, **required for Pro `first_last_frame` mode**
</ParamField>

<ParamField body="last_frame_image" type="string">
  Last-frame image URL, **required for Pro `first_last_frame` mode**
</ParamField>

<ParamField body="callback_url" type="string">
  Webhook callback URL, invoked when the task reaches a terminal state. See [Callback Mechanism](/en/docs/sync-async#async-task-webhook)
</ParamField>

<ParamField body="callback_events" type="string[]">
  Subscribed callback events; defaults to terminal events (`succeeded` / `failed` / `cancelled`)
</ParamField>

## Model-specific Parameters

<Tabs>
  <Tab title="viduq3-pro">
    <ParamField body="audio" type="boolean" default="true">
      Whether to generate audio (default `true`). Pro tier enables **audio output** by default (dialog + SFX); set `false` for silent output
    </ParamField>

    Pro tier auto-routes the generation mode by `image_urls` count; see [Auto-routing](#auto-routing).
  </Tab>

  <Tab title="viduq3-mix">
    Mix tier has no specific parameters. Note that `resolution` only supports `720P` / `1080P`, **540P is not supported**.

    Mix tier enables **multi-asset joint generation** by default — pass multiple references via `image_urls` (1-7) and the system applies smart transitions, ideal for short-drama re-cutting and ad creative remix.
  </Tab>

  <Tab title="viduq3">
    Standard tier has no specific parameters; all options live in [Common Parameters](#common-parameters).

    The default daily choice, with smarter shot transitions; accepts 1-7 images as subject / style references.
  </Tab>

  <Tab title="viduq3-turbo">
    Turbo tier has no specific parameters.

    <Note>
      This model may be unavailable in some scenarios; please contact sales if needed.
    </Note>
  </Tab>
</Tabs>

## Auto-routing

`viduq3-pro` auto-selects the generation mode by `image_urls` array length, so **you usually don't need to pass `action` explicitly**:

| `image_urls` count | Routed mode          | Equivalent `action` | Notes                                                |
| ------------------ | -------------------- | ------------------- | ---------------------------------------------------- |
| `0` (or omitted)   | **Text-to-video**    | `generate`          | Pure prompt-driven                                   |
| `1`                | **Image-to-video**   | `image2video`       | The image serves as the starting frame               |
| `2`                | **First-last frame** | `first_last_frame`  | First image = first frame, second image = last frame |

<Note>
  When `first_frame_image` + `last_frame_image` are passed explicitly, they take precedence and override the first-last-frame routing from `image_urls`.
</Note>

Other tiers (`viduq3` / `viduq3-mix` / `viduq3-turbo`) default to **R2V reference-to-video** mode, with `image_urls` providing 1-7 subject / style references.

## Resource Limits

| Item                    | Limit                                                     |
| ----------------------- | --------------------------------------------------------- |
| Reference images        | Up to 7, each ≤ 30MB, JPG / PNG / WEBP                    |
| First/last-frame images | JPG / PNG / WEBP, public download URL required            |
| Prompt length           | ≤ 2000 characters recommended                             |
| Video duration          | Any integer 1-16 sec (Mix supports 1-second short videos) |
| Output                  | MP4, link valid for 24 hours                              |

## Related Docs

* [Task System Reference](/en/docs/task-system) — task state machine / polling cadence / async push
* [Request & Response Format](/en/docs/request-response) — common error codes / headers / rate limits
* [Authentication](/en/docs/authentication) — API key signup and usage