Prohibited content and uses
- Pornography and NSFW: Generating or distributing pornographic, explicit, sexually suggestive, or sexually exploitative content.
- Violence and gore: Encouraging real-world harm or crime, or creating extreme graphic violence or abuse.
- Hate: Inciting hatred, discrimination, or violence against protected groups.
- Child endangerment: Any sexualization, sexual exploitation, abuse, or grooming of minors, including fictional and synthetic content.
- Deepfakes and impersonation: Using the service for face-swapping, deepfake, or facial manipulation tools is prohibited. Deceptive identity content, non-consensual intimate imagery, and impersonation for fraud are prohibited.
- Copyright and trademark infringement: Using assets without the necessary rights, misusing brands, or creating content that infringes others' rights.
Do not bypass safety restrictions, abuse accounts, attack the service, falsify payments or bills, or use batch tasks to carry out these activities.
Models and safety restrictions
Online generation is subject to the selected model's and provider's safety restrictions. Requests or results may be refused and tasks may fail; enforcement mechanisms differ by service. Do not evade restrictions by changing channels, repeatedly rewriting prompts, or using multiple accounts.
This policy covers all available models. Provider filtering and our manual review of reports serve complementary roles; we do not promise that every input or output has been individually reviewed by a person. Suspected violations may be reviewed using provider responses, reports, and necessary task records, with service restrictions or other measures applied as permitted by law.
Report a violation
Please use the reporting channel to submit the issue type, relevant task ID, necessary screenshots or accessible links, and your relationship to the rights in the material. Do not distribute complete unlawful content or others' sensitive information to demonstrate the issue.
We review the account, task, and service scope involved in the report. Mark urgent safety issues clearly; if someone is in danger, also contact the appropriate emergency authorities.
Enforcement and appeals
Depending on the facts and severity, action may include refusing requests, restricting service, suspending relevant accounts, or requesting correction of publicly displayed content. We cannot remotely delete projects stored only on a user's device. Content held by third parties must be addressed through their respective channels.
Balances and ledgers are handled according to transaction records and applicable rules; a report does not automatically forfeit all usage. If you believe a decision is incorrect, contact support and provide task or ticket details to request a review.