Content Policy

The rules are checked before anything generates, so a blocked request never costs a credit and never pretends to work. Here's what's blocked and why.

Never allowed

• Anything sexualizing minors — absolute block, logged, zero tolerance.
• Sexual or pornographic content — the underlying image models refuse it upstream; we tell you honestly instead of failing silently.
• Real people — no generating identifiable real people, and no sexualized or compromising depictions of anyone.
• Graphic violence and gore.

How enforcement works

• Requests are screened twice: a fast rules check, then an AI classification — both before any credit is spent.
• If the model itself refuses a combination, you'll see a plain message asking you to reword; nothing is charged.
• Blocks are logged. Repeated deliberate attempts to evade the rules close the account.

Edge cases, honestly

• Some innocent words trip the fast filter (medical terms, food photography). If a reasonable request is blocked, reword it — the second-tier check understands context better.
• Cartoon/stylized characters follow the same rules as photorealistic ones.
• Own-brand watermark or text removal on your own images is fine; using it on others' work isn't, and usage is logged.

Reporting

If the service produces something that shouldn't exist, use the report option on the image — every image carries a full generation trace, so reports are actionable immediately.