Content and usage responsibility

Model capabilities and usage

Choose a model for the capability you need. SpicyAPI does not require a content-mode request flag and does not perform per-request content review.

In plain terms: you choose the model; the model decides what it will generate. We don't scan, classify or approve individual requests, and we don't rewrite your prompt. There is no "adult mode" switch to turn on and no approval to apply for — what a model can do is part of that model, and its catalog page describes it.

Your responsibility

API execution does not determine whether content is suitable for distribution in your jurisdiction and does not replace your compliance process. You are responsible for making the use, input, output, audience and storage comply with applicable law, contracts and your product policy, especially where real people, consent, age assurance or protection of minors are involved.

If you let your own users send requests through your product, these duties stay with you: age checks your audience requires, your own content and access controls, and a way for your users to report content to you.

What is never allowed

No per-request review doesn't mean anything goes. Some uses are prohibited on every model, with no setting, account type or agreement that changes it, including:

• any sexualised depiction of a minor, in any style;
• intimate or sexualised depictions of an identifiable real person without their documented consent, including face-swap and likeness transfer;
• impersonating a real person;
• content that is unlawful where you are or where the service is operated.

Accounts that break these rules can be suspended or closed after a report or legal notice, and material involved in a report is preserved where the law requires it. Read the Acceptable Use Policy and the Content Policy before you launch, especially if you serve your own end users.

Where we operate

We do not serve mainland China. This is a legal compliance requirement, not a product decision.

The check is enforced at the edge, at the origin, at sign-up and at payment. Requests from a restricted region are refused:

• Most get business code 40303 (The service is not available in your region).
• Some are stopped by our edge network before they reach the API and get a plain 403 page instead of a JSON response.
• A request whose location can't be determined is refused the same way.

A 40303 can't be retried or worked around.

Separately, the terms of service do not offer the service in the United Kingdom, Hong Kong, Macau, Saudi Arabia, the United Arab Emirates, Qatar, Kuwait, Bahrain, Oman, Pakistan, Bangladesh, Afghanistan, Indonesia, Malaysia or Brunei, because of how the law there treats adult and AI-generated content. Using the service from those places is not permitted under the terms.

Model rejection

A specific model may reject a request according to its implementation. The task ends as failed with errorCode: "content_rejected", and the hold is normally released in full. Some content refusals are charged at the accepted quote only when the model page explicitly states this. These remain failed tasks; settled: true with a positive cost confirms a charge. A rejection is not automatically chargeable.

errorCode is stable; errorMessage is for people to read and may quote the model's own reason in English, with internal details removed (see Errors). Branch on errorCode, and don't retry a rejection unchanged: change the prompt, or choose a different model.

Reporting

To report abuse, email support@spicyapi.ai or use the reporting page. Include the task ID (it starts with job_ and is shown in the console under Logs) or the URL of the page where the content appears. Do not send API keys or signed download URLs.