> ## 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.

# Veo Series

> Google Veo 3.1 — official direct Fast / Quality tiers, reverse-engineered Lite / Quality tiers, native audio, first/last frame, Remix

Google Veo 3.1 video generation, available through both **official direct** and **reverse-engineered** channels, with **4 variants** in total.

* **Official tiers** (Vertex AI direct): billed **per second**; supports 4 / 6 / 8 second durations and 720P / 1080P / 4K resolutions.
* **Reverse-engineered tiers**: billed **per call** (fixed price per call); only **8-second** fixed duration.

| Variant                   | Channel            | Billing    | Positioning                                 |
| ------------------------- | ------------------ | ---------- | ------------------------------------------- |
| `veo3.1-quality-official` | Official direct    | Per-second | Flagship tier, final-delivery quality       |
| `veo3.1-fast-official`    | Official direct    | Per-second | Fast tier, suited for iteration drafts      |
| `veo3.1-quality`          | Reverse-engineered | Per-call   | Reverse-engineered flagship, fixed 8 sec    |
| `veo3.1-lite`             | Reverse-engineered | Per-call   | Reverse-engineered lightweight, fixed 8 sec |

Both channels natively support audio, first/last-frame keyframes, and Remix continuation.

## Pricing

<Note>
  **4K resolution does not support muted output** — you must set `generate_audio=true` (720p / 1080p can be either muted or with audio). In the table below, `720p_sound` / `1080p_sound` are the audio-on premium tiers, and `4k_sound` is the only available 4K tier.
</Note>

### Official Tiers (per-second)

720P and 1080P share the same price (Google's official tiered pricing); enabling `generate_audio=true` **doubles** the unit price.

#### `veo3.1-fast-official`

| Spec          | Unit price (per sec) | 8-second video |
| ------------- | -------------------- | -------------- |
| 720P muted    | `$0.08`              | `$0.64`        |
| 1080P muted   | `$0.08`              | `$0.64`        |
| 720P + audio  | `$0.12`              | `$0.96`        |
| 1080P + audio | `$0.12`              | `$0.96`        |
| 4K + audio    | `$0.28`              | `$2.24`        |

#### `veo3.1-quality-official`

| Spec          | Unit price (per sec) | 8-second video |
| ------------- | -------------------- | -------------- |
| 720P muted    | `$0.16`              | `$1.28`        |
| 1080P muted   | `$0.16`              | `$1.28`        |
| 720P + audio  | `$0.32`              | `$2.56`        |
| 1080P + audio | `$0.32`              | `$2.56`        |
| 4K + audio    | `$0.48`              | `$3.84`        |

### Reverse-Engineered Tiers (per-call)

Reverse-engineered tiers use a **fixed per-call price** independent of duration, resolution, or audio (only 8 seconds is supported). Refer to the admin panel / console for the published price.

| Model ID         | Billing  | Duration          | Per-call price    |
| ---------------- | -------- | ----------------- | ----------------- |
| `veo3.1-quality` | Per-call | 8 seconds (fixed) | TBD (see console) |
| `veo3.1-lite`    | Per-call | 8 seconds (fixed) | TBD (see console) |

## Request Examples

<CodeGroup>
  ```bash Text-to-video (minimal) theme={"system"}
  curl -X POST https://www.qingbo.dev/v1/tasks \
    -H "Authorization: Bearer $WAVE_API_KEY" \
    -H "Content-Type: application/json" \
    -d '{
      "model": "veo3.1-fast-official",
      "prompt": "Neon-lit street on a rainy night, a red sports car drives through a puddle, cinematic slow motion",
      "duration": 8,
      "resolution": "1080p",
      "aspect_ratio": "16:9"
    }'
  ```

  ```bash Image-to-video (I2V first frame) theme={"system"}
  curl -X POST https://www.qingbo.dev/v1/tasks \
    -H "Authorization: Bearer $WAVE_API_KEY" \
    -H "Content-Type: application/json" \
    -d '{
      "model": "veo3.1-quality-official",
      "action": "image2video",
      "prompt": "The person slowly looks up and smiles, hair blown by the wind",
      "first_frame_image": "https://cdn.example.com/portrait.jpg",
      "duration": 6,
      "resolution": "1080p",
      "resize_mode": "pad"
    }'
  ```

  ```bash With native audio (generate_audio=true) theme={"system"}
  curl -X POST https://www.qingbo.dev/v1/tasks \
    -H "Authorization: Bearer $WAVE_API_KEY" \
    -H "Content-Type: application/json" \
    -d '{
      "model": "veo3.1-quality-official",
      "prompt": "Waves crashing on rocks, seagulls calling, a distant ship horn",
      "duration": 8,
      "resolution": "1080p",
      "aspect_ratio": "16:9",
      "generate_audio": true
    }'
  ```

  ```bash Remix continuation theme={"system"}
  curl -X POST https://www.qingbo.dev/v1/tasks \
    -H "Authorization: Bearer $WAVE_API_KEY" \
    -H "Content-Type: application/json" \
    -d '{
      "model": "veo3.1-fast-official",
      "action": "remix",
      "prompt": "Camera continues to push in, the figure turns and walks away",
      "first_frame_image": "https://cdn.example.com/last-frame.jpg",
      "duration": 4,
      "resolution": "1080p",
      "raw": true
    }'
  ```

  ```bash Reverse-engineered tier (fixed 8 sec) theme={"system"}
  curl -X POST https://www.qingbo.dev/v1/tasks \
    -H "Authorization: Bearer $WAVE_API_KEY" \
    -H "Content-Type: application/json" \
    -d '{
      "model": "veo3.1-quality",
      "prompt": "A seaside town at sunset, doves flying past the bell tower",
      "duration": 8,
      "aspect_ratio": "16:9"
    }'
  ```
</CodeGroup>

<ResponseExample>
  ```json Submit success theme={"system"}
  {
    "task_id": "task-wave1775285160b950328499",
    "model": "veo3.1-quality-official",
    "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                  | Channel            | Variant                  | Billing    | Duration      | Resolution        |
| ------------------------- | ------------------ | ------------------------ | ---------- | ------------- | ----------------- |
| `veo3.1-quality-official` | Official direct    | Quality flagship         | Per-second | 4 / 6 / 8 sec | 720P / 1080P / 4K |
| `veo3.1-fast-official`    | Official direct    | Fast tier                | Per-second | 4 / 6 / 8 sec | 720P / 1080P / 4K |
| `veo3.1-quality`          | Reverse-engineered | Quality flagship variant | Per-call   | 8 sec (fixed) | 720P / 1080P      |
| `veo3.1-lite`             | Reverse-engineered | Lite lightweight variant | Per-call   | 8 sec (fixed) | 720P / 1080P      |

## Common Parameters

<ParamField body="model" type="string" required>
  Model ID — one of `veo3.1-quality-official` / `veo3.1-fast-official` / `veo3.1-quality` / `veo3.1-lite`
</ParamField>

<ParamField body="action" type="string" default="generate">
  Operation type, valid values:

  * `generate` — text-to-video (T2V)
  * `image2video` — image-to-video (I2V); requires `first_frame_image`
  * `first_last_frame` — first/last-frame keyframes; requires both `first_frame_image` and `last_frame_image`
  * `remix` — continuation based on an existing clip
</ParamField>

<ParamField body="prompt" type="string" required>
  Video description text; required for all actions
</ParamField>

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

  * `16:9` — landscape widescreen
  * `9:16` — portrait
</ParamField>

<ParamField body="resolution" type="string" default="1080p">
  Output resolution, valid values:

  * `720p`
  * `1080p`
  * `4k` — **official tiers only**, and **must** have `generate_audio=true` (4K does not support muted output)
</ParamField>

<ParamField body="duration" type="integer" default="8">
  Video duration in seconds:

  * **Official tiers**: `4` / `6` / `8`
  * **Reverse-engineered tiers**: `8` only (fixed)
</ParamField>

<ParamField body="image_urls" type="string[]">
  Generic image-reference array (in some scenarios equivalent to `first_frame_image`)
</ParamField>

<ParamField body="first_frame_image" type="string">
  First-frame image URL. Required for `image2video` / `first_last_frame` / `remix` modes
</ParamField>

<ParamField body="last_frame_image" type="string">
  Last-frame image URL — **required only in `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 event types; defaults to terminal states (`completed` / `failed`)
</ParamField>

## Model-Specific Parameters

<Note>
  **All four variants share the same parameters**; only the channel, billing model, and available duration / resolution differ. Tab order is Quality before Fast / Lite, and official before reverse-engineered.
</Note>

<Tabs>
  <Tab title="veo3.1-quality-official">
    <ParamField body="generate_audio" type="boolean" default="false">
      Whether to generate the audio track. When enabled, unit price doubles; **4K must** be `true`
    </ParamField>

    <ParamField body="sample_count" type="integer" default="1">
      Number of samples to generate; range 1-4
    </ParamField>

    <ParamField body="resize_mode" type="string" default="pad">
      Image resize strategy for image-to-video, valid values:

      * `pad` — edge padding, preserves the entire frame
      * `crop` — center crop, fits the output ratio
    </ParamField>

    <ParamField body="enhance_prompt" type="boolean" default="true">
      Upstream prompt enhancement; **must be `true`** (forced on by Vertex AI)
    </ParamField>

    <ParamField body="person_generation" type="string" default="allow_adult">
      Person-generation policy, valid values:

      * `allow_adult` — allow generating adult persons
      * `disallow` — disallow generating persons
    </ParamField>

    <ParamField body="raw" type="boolean" default="false">
      Remix-mode only; when `true`, returns only the extension, not concatenated with the original clip
    </ParamField>
  </Tab>

  <Tab title="veo3.1-fast-official">
    <ParamField body="generate_audio" type="boolean" default="false">
      Whether to generate the audio track. When enabled, unit price doubles; **4K must** be `true`
    </ParamField>

    <ParamField body="sample_count" type="integer" default="1">
      Number of samples to generate; range 1-4
    </ParamField>

    <ParamField body="resize_mode" type="string" default="pad">
      Image resize strategy for image-to-video, valid values:

      * `pad` — edge padding, preserves the entire frame
      * `crop` — center crop, fits the output ratio
    </ParamField>

    <ParamField body="enhance_prompt" type="boolean" default="true">
      Upstream prompt enhancement; **must be `true`** (forced on by Vertex AI)
    </ParamField>

    <ParamField body="person_generation" type="string" default="allow_adult">
      Person-generation policy, valid values:

      * `allow_adult` — allow generating adult persons
      * `disallow` — disallow generating persons
    </ParamField>

    <ParamField body="raw" type="boolean" default="false">
      Remix-mode only; when `true`, returns only the extension, not concatenated with the original clip
    </ParamField>
  </Tab>

  <Tab title="veo3.1-quality (reverse)">
    <ParamField body="generate_audio" type="boolean" default="false">
      Whether to generate the audio track
    </ParamField>

    <ParamField body="sample_count" type="integer" default="1">
      Number of samples to generate; range 1-4
    </ParamField>

    <ParamField body="resize_mode" type="string" default="pad">
      Image resize strategy for image-to-video, valid values:

      * `pad` — edge padding, preserves the entire frame
      * `crop` — center crop, fits the output ratio
    </ParamField>

    <ParamField body="person_generation" type="string" default="allow_adult">
      Person-generation policy, valid values:

      * `allow_adult` — allow generating adult persons
      * `disallow` — disallow generating persons
    </ParamField>

    <ParamField body="raw" type="boolean" default="false">
      Remix-mode only; when `true`, returns only the extension, not concatenated with the original clip
    </ParamField>

    <Note>
      Reverse-engineered tiers are **billed per call**; `duration` only accepts `8`, and `resolution` does not support `4k`.
    </Note>
  </Tab>

  <Tab title="veo3.1-lite (reverse)">
    <ParamField body="generate_audio" type="boolean" default="false">
      Whether to generate the audio track
    </ParamField>

    <ParamField body="sample_count" type="integer" default="1">
      Number of samples to generate; range 1-4
    </ParamField>

    <ParamField body="resize_mode" type="string" default="pad">
      Image resize strategy for image-to-video, valid values:

      * `pad` — edge padding, preserves the entire frame
      * `crop` — center crop, fits the output ratio
    </ParamField>

    <ParamField body="person_generation" type="string" default="allow_adult">
      Person-generation policy, valid values:

      * `allow_adult` — allow generating adult persons
      * `disallow` — disallow generating persons
    </ParamField>

    <ParamField body="raw" type="boolean" default="false">
      Remix-mode only; when `true`, returns only the extension, not concatenated with the original clip
    </ParamField>

    <Note>
      Reverse-engineered tiers are **billed per call**; `duration` only accepts `8`, and `resolution` does not support `4k`.
    </Note>
  </Tab>
</Tabs>

## Resource Limits

| Item            | Limit                                                                                 |
| --------------- | ------------------------------------------------------------------------------------- |
| Reference image | JPG / PNG / WEBP, ≤ 20MB each, 720P or higher recommended                             |
| Output video    | MP4, link valid for 24 hours                                                          |
| Duration        | Official tiers 4 / 6 / 8 sec; reverse-engineered tiers fixed 8 sec only               |
| Resolution      | 720P / 1080P freely; **4K only on official tiers** and requires `generate_audio=true` |
| Audio           | Natively generated, no extra TTS needed; 4K must include audio, 720P / 1080P optional |
| Persons         | Constrained by `person_generation`; non-compliant content is blocked upstream         |

## Related Docs

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