Content Policy

Last updated: 2026-04-23

Image-generation tools can be misused. This policy describes what users may not create, upload, or distribute using Renderbench. It applies to every prompt, every input image, and every generated output on the Service, and is incorporated by reference into our Terms of Service.

Absolutely prohibited

The following are banned on Renderbench. Attempts to generate, upload, or share content in any of these categories will result in immediate account termination and, where required by law, a report to the appropriate authorities:

• Child sexual abuse material, and any content that sexualizes, grooms, or exploits minors in any form. This is an absolute prohibition with zero tolerance.
• Non-consensual intimate imagery, including nude or sexual deepfakes of real, identifiable people.
• Real-person sexual content generated, altered, or distributed without that person’s explicit consent.
• Content inciting or glorifying violence, terrorism, self-harm, or mass casualty events.
• Doxxing content — generating imagery designed to expose private information about real, identifiable individuals (home addresses, workplaces, medical status).
• Election and civic disinformation — content that falsely depicts real political figures or events with the intent to mislead voters during an active election cycle.

Restricted — use at your own legal risk

The following aren’t universally prohibited but can create legal exposure for you. You must have all necessary rights, releases, and lawful basis before using Renderbench for any of these:

• Generating content depicting real, identifiable people, including celebrities — the subject’s rights of publicity and privacy still apply.
• Uploading third-party imagery as references or inputs — you must hold the copyright, have a license, or fall under a valid exception (fair use, etc.).
• Generating content in the style of a specific living artist — this may implicate that artist’s rights under your jurisdiction.
• Commercial use of generated outputs — subject to the upstream model provider’s terms (for example, OpenRouter’s model-routing rules and the downstream provider’s commercial-use policy).

Upstream model policies

Renderbench sends your prompts and images to third-party AI models (including Google Gemini and Black Forest Labs FLUX, routed via OpenRouter). Those providers apply their own safety filters and content policies in addition to ours. A prompt that is allowed here may still be rejected by the upstream model. If a generation fails with a safety-related error, the failure is reflected in the Service but the details come from the provider.

Moderation

We rely on a combination of provider-side safety filters, account-level behaviour signals, and user reports. We do not proactively review the content of every prompt or output, but we review content promptly on receipt of an abuse report or if automated signals indicate a violation.

Reporting abuse

If you encounter misuse of the Service, email hello@renderbench.com with the subject line “Abuse Report” and include links or identifiers (project, experiment, or version IDs where possible). We triage abuse reports within one business day. For emergencies involving imminent danger to a child or other person, also contact local authorities — that’s not something we can substitute for.

Enforcement

Depending on the severity and your history, enforcement actions may include:

• Removing the specific content.
• Sending a warning and requiring you to acknowledge this policy.
• Suspending your account.
• Permanently terminating your account and blocking your payment method.
• Reporting to law enforcement or relevant authorities where required by law.

We don’t grade on a curve — one instance of the “Absolutely prohibited” categories above is enough to lose access to the Service permanently.

Appeals

If you believe an enforcement action was mistaken, reply to the enforcement email within 30 days explaining why. We’ll review and respond within a reasonable timeframe.

Contact

Abuse reports and policy questions: hello@renderbench.com.