Policy

Tevyon is a tool for creatives, brands and developers. This policy applies to all inputs and results, through every access point: Studio, API, MCP, command line and connectors. It is part of the Terms. The makers of the models have their own rules. Whatever one of their models refuses, Tevyon refuses too.

What counts is the effect of a piece of content, not how the prompt is worded. Prohibited content stays prohibited, even if it is meant as a joke, art, a test or “just for me”.

• Sexual or sexualized depictions of minors, or of characters who appear to be minors, are prohibited. This applies to every style, including drawings and animations.
• Content that endangers or exposes minors, or lures them into contact with adults, is prohibited.
• You may only upload images of real minors if you have parental authority over them and the use is harmless.
• For material showing the sexual abuse of children, we suspend the account immediately and report it to the competent authorities.

The following applies to the face, body, voice and name of real people:

• You need the consent of the person shown if you use their image, their voice or their identity as input. For public figures, satire and commentary are allowed if it is clearly recognizable that it is fiction.
• Sexual or intimate depictions of real people without their consent are prohibited, even if they are made up.
• Content that convincingly makes a real person say or do something they did not say or do is prohibited if it can mislead viewers. This applies especially to politics, elections, news, the justice system and health.
• Harassment, bullying, blackmail and damaging someone’s reputation with content about real people are prohibited.
• Identifying people by their face or extracting biometric features from inputs is prohibited.

If you publish realistic depictions of real people, places or events, you must label them as AI-generated. How to do that is explained under AI Labeling.

• Sexual content: Pornography and explicit nudity are currently not allowed on Tevyon.
• Violence: glorification of real violence, instructions for violent acts, depictions of real victims for amusement.
• Terrorism and extremism: propaganda, recruitment or glorification of terrorist or violent extremist groups.
• Hate: degrading people, or calling for violence against them, because of their origin, religion, gender, sexual orientation, disability or similar characteristics.
• Self-harm: content that promotes suicide, self-harm or eating disorders.
• Fraud and deception: fake IDs, receipts, certificates, reviews or screenshots. Phishing, identity theft and content that imitates authorities, companies or people in order to deceive.
• Disinformation: deliberately misleading content about elections, referendums, crises or public health.
• Dangerous goods: instructions for weapons, explosives or drugs. Advertising for illegal goods or services.

• Only upload images, videos, music and texts that you hold the rights to or have permission to use.
• Don’t create results that reproduce protected characters, logos or works so closely that they can be confused with the original, and don’t use them commercially without permission.
• Don’t use other people’s trademarks in a way that creates the impression of a partnership or of commercial origin.
• Don’t remove watermarks or provenance information from third-party works, or AI labels from results.

• Don’t circumvent moderation, filters or blocks, not even with encoded, split-up or step-by-step prompts.
• Don’t use automated access outside the official API, the MCP server and the command line, and stay within their limits.
• Don’t share accounts or keys outside your workspace, and don’t resell access.
• Don’t use results or responses from the service to train a competing AI model if the model’s maker prohibits this.
• Don’t attack the security of the service. Please report any vulnerabilities you find to support@tevyon.ai; we will thank you for it.

Inputs and results are checked automatically, by us and by the makers of the models. If the check blocks a job, you get your credits back and are told that the content violates the rules. Repeated or serious violations are reviewed by people.

Depending on the severity, we respond in stages:

1. a notice and removal of the content,
2. restriction of individual features or models,
3. temporary suspension of the account,
4. termination of the account. For child abuse material or serious threats, this happens immediately, and we inform the authorities.

We give reasons for every measure. You can appeal suspensions and terminations within 30 days to support@tevyon.ai. The appeal is reviewed by a person who was not involved in the first decision.

We accept reports at support@tevyon.ai, even without an account. A report should include: where the content can be found, why it is unlawful or violates the rules, your name and your email address, except for child abuse material, and a statement that your information is correct to the best of your knowledge.

We confirm receipt, review carefully and tell you our decision. We remove intimate images of a person that were made without their consent within 48 hours of a substantiated report. Rights holders can report copyright infringements the same way.

In the Community, you report a creation with “Report” in its large view and a profile with “Report” on the profile page, each with a reason and an optional note; you need an account for this. Whoever is reported won’t find out who reported them. After several reports, or after a single one if minors are involved, we take a creation out of the Community for the time being until a person has reviewed it. If we take out a creation or a profile, the person sees the reason in their profile.