> ## Documentation Index
> Fetch the complete documentation index at: https://docs.ruapi.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Seedance: PolicyViolation and SensitiveContentDetected — why a prompt, photo or finished video was rejected

> The Seedance 2.0 / 2.5 errors “InputImageSensitiveContentDetected.PolicyViolation” (“may be related to copyright restrictions”), “InputTextSensitiveContentDetected” and “OutputVideoSensitiveContentDetected” (“may contain sensitive information”): what the model checks, why celebrities and characters are rejected, what to do and whether you're charged.

**In short:** these are moderation errors from the Seedance model itself (BytePlus ModelArk). The model checks both what you send (text, images, video, audio) and what comes out. `PolicyViolation` means a possible copyright issue: most often the input contains a celebrity or a character from a film, TV series or game. `SensitiveContentDetected` with no suffix means prohibited content in the prompt, the input files or the finished video. Only changing the input helps; you're not charged for a rejected task.

If the error says *may contain real person* (`PrivacyInformation`), that's a different case — see [Seedance: input image may contain real person](https://docs.ruapi.ai/en/seedance-real-person-error.md).

## Error codes

In ModelArk, all of them have HTTP status **400**. The error messages come from the [ModelArk error code table](https://docs.byteplus.com/en/docs/ModelArk/1299023):

| Code | Error message | What was checked |
| - | - | - |
| `InputTextSensitiveContentDetected` | The request failed because the input text may contain sensitive information. | Prompt |
| `InputImageSensitiveContentDetected` | The request failed because the input image may contain sensitive information. | Input image |
| `InputVideoSensitiveContentDetected` | The request failed because the input video may contain sensitive information. | Reference video |
| `InputAudioSensitiveContentDetected` | The request failed because the input audio may contain sensitive information. | Input audio |
| `OutputVideoSensitiveContentDetected` | The request failed because the output video may contain sensitive information. | Finished video |
| `InputTextSensitiveContentDetected.PolicyViolation` | The request failed because the input text may be related to copyright restrictions. | Prompt: a character's or celebrity's name |
| `InputImageSensitiveContentDetected.PolicyViolation` | The request failed because the input image may be related to copyright restrictions. | Image: a celebrity or a character |
| `InputVideoSensitiveContentDetected.PolicyViolation` | The request failed because the input video may be related to copyright restrictions. | Reference video |
| `InputAudioSensitiveContentDetected.PolicyViolation` | The request failed because the input audio may be related to copyright restrictions. | Input audio |
| `OutputVideoSensitiveContentDetected.PolicyViolation` | The request failed because the output video may be related to copyright restrictions. | Finished video |

The message usually ends with `Request ID: ...`.

On RuAPI the error arrives in one of two ways:

* **Right at submission.** If the reason is copyright or a public figure, you get a `400` response with the code `input_copyright_restricted` along with a hint about what to do. For other moderation rejections the response is currently generic — `upstream task submission failed (HTTP …)` — and doesn't show which code from the table was triggered. Check your prompt and files against the "What you can do" table below.
* **During generation.** The task ends with the status `failed`, with one of the codes from the table in the `error` field.

Either way, the prepayment is refunded to your balance in full.

## Why it happens

* **Public figures.** The model doesn't accept photos of celebrities, influencers or streamers — and `auto_create_assets` doesn't help here.
* **Copyrighted characters.** Characters from films, TV series, cartoons and games — whether they appear in a picture or are only named in the prompt.
* **Prohibited content.** Violence, explicit scenes and other topics banned by the model's rules — in the prompt, the input files or the output.
* **The check errs on the side of caution.** The error says *may*. Sometimes a perfectly harmless request is rejected too.

## What you can do

| Error | What to do |
| - | - |
| `...PolicyViolation` on input | Remove names of celebrities and characters, and titles of films and games, from the prompt. Replace the image with your own — an original character or an ordinary person (photos of ordinary people usually work; some mini and fast modes need `auto_create_assets`, see [Seedance: real person](https://docs.ruapi.ai/en/seedance-real-person-error.md)) |
| `InputTextSensitiveContentDetected` | Rewrite the prompt in more neutral terms: drop harsh words and descriptions of violence or explicit scenes |
| `InputImage/Video/AudioSensitiveContentDetected` | Replace the input file |
| `OutputVideoSensitiveContentDetected.PolicyViolation` | The finished video looks like protected content. Remove references to well-known characters, films and brands from the prompt and submit the task again |
| `OutputVideoSensitiveContentDetected` | The finished video failed the check. Tone down the prompt or change the scene and submit the task again |

If your project really needs a public figure and you have their consent, contact support at [support@ruapi.ai](mailto:support@ruapi.ai) or on Telegram [@RuAPI\_Support](https://t.me/RuAPI_Support).

## How to use Seedance via RuAPI

**1. Sign up**

Sign up at [www.ruapi.ai](https://www.ruapi.ai/?lng=en\&utm_source=docs\&utm_campaign=en-seedance-sensitive-content-error) — with email, Google or Yandex.

**2. Top up**

Console → **Wallet**: USDT or USDC (TRC20 / BEP20 / Polygon; USDT also on TON), from 5 USDT. No USDT yet? [Buy it with rubles via SBP (Faster Payments System)](https://docs.ruapi.ai/en/buy-usdt.md). Cards from foreign banks (Visa / Mastercard) are accepted too. Video models become available after your first top-up.

**3. Key**

Console → **API Keys** → **Create API key** — you'll get an `sk-...` string.

**4. Request**

Examples for every mode are in the [video generation guide](https://docs.ruapi.ai/en/video-generation.md). If your code is already written for ModelArk, change the base URL to `https://www.ruapi.ai/v1`, and swap in your RuAPI key and our model name (for example, `doubao-seedance-2.5`).

## FAQ

> [accordion] Am I charged for a rejected task?
>
> No. Only successfully generated videos are billed: if a task is rejected at submission or ends with an error — including `OutputVideoSensitiveContentDetected` — the prepayment is refunded to your balance in full.

> [accordion] Will resending the same request help?
>
> For input errors, no: the same text or file will be rejected again. With `OutputVideoSensitiveContentDetected` you can try again, but changing the prompt is more reliable.

> [accordion] Can the check be turned off?
>
> No. It's the Seedance model's own check and can't be disabled.

> [accordion] Do I need a BytePlus account or a VPN?
>
> No. A RuAPI key is enough: the website and the API work without a VPN.

## Next

> [card] [Video generation: Seedance 2.0 / 2.5](https://docs.ruapi.ai/en/video-generation.md)
>
> All modes, the ModelArk format, draft mode and prices.

> [card] [Seedance: real person](https://docs.ruapi.ai/en/seedance-real-person-error.md)
>
> Photos and videos of real people, auto\_create\_assets.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.