Acceptable Use Policy

Last updated: 13 October 2026

Uncensored does not mean anything goes. The models on unrestricted.sh are built to help with legitimate work, such as coding, research, security testing, and writing, instead of refusing it. They are not a way around the law or around this policy. This policy sets out what is never allowed. It is part of our Terms of Service, and it applies to the chat playground, the API, and anything you build on top of them.

Never allowed

Do not use the service to create, request, or distribute:

Sexual content involving minors. This includes any content that sexualises a person under 18, whether real or fictional. Accounts that break this rule are closed and reported to the authorities when we become aware of it.
Sexually explicit content, including erotic or pornographic writing and roleplay. Mature themes in fiction are fine; explicit sexual material is not. This rule applies from 21 October 2026.
Intimate content of real people without their consent, including sexual deepfakes of real people.
Help with violence against real people, such as planning an attack, identifying targets, or threatening someone.
Terrorism or violent extremism, including recruitment, propaganda meant to incite violence, or operational help.
Weapons capable of mass casualties, meaning biological, chemical, nuclear, or radiological weapons.
Attacks on computer systems you do not have permission to test, including malware meant for use against others, intrusions, and denial-of-service attacks.
Fraud and deception, including phishing, scams, fake reviews, forged documents, and impersonating real people or organisations to mislead others.
Harassment, stalking, or doxxing, including collecting or publishing private information about a person.
Large-scale spam or manipulation, such as bulk unsolicited messages or coordinated campaigns that present generated content as coming from many real people.
Anything illegal under Danish or EU law, or under the law that applies where you are.

Writing fiction within the rules above, doing research, testing security with permission, and discussing difficult topics are allowed. What matters is whether the use harms, or is meant to harm, real people or systems.

Keeping the service fair

One account per person. Do not create several accounts, for example to collect welcome credit more than once.
Do not share your account or give your API keys to people outside your own organisation.
Do not resell direct access to the API, such as a proxy that passes other people's requests to it, without our written permission. You may build and sell your own products that use the API.
Do not try to get around limits, billing, or suspensions, and do not probe, overload, or attack the service.

If you build on unrestricted.sh

You are responsible for your own users following this policy. Put safeguards in place that fit your product and your audience, and do not offer the models to people under 18. You are also responsible for telling your users that they are interacting with an AI system, and for labelling AI-generated content where the law requires it, such as under Article 50 of the EU AI Act.

Enforcement

We do not save or log prompts or outputs. We may use automated checks that run in memory while a request is handled, to detect and block sexual content involving minors; these checks do not save prompts or outputs. We also act on usage patterns, payment and account signals, reports from others, and requests from authorities. If you break this policy, we may:

remove welcome credit,
suspend or close your account, as described in the Terms of Service, and
report the matter to the authorities where the law requires it or where people are at risk.

Report abuse

To report misuse of unrestricted.sh, email abuse@unrestricted.sh with as much detail as you can. If you get no answer, write to support@unrestricted.sh.