Acceptable Use Policy

1. Principles

Use the service lawfully, do not harm people, do not harm the platform, and respect the usage policies of the model vendors your requests reach. You are responsible for the content you send, for what your applications let end users send, and for how outputs are used.

2. Prohibited content and uses

Do not use ElevenRouter to create, request, store or distribute content or to run activity that:

• is illegal where you or your end users are, or infringes intellectual-property, privacy or publicity rights;
• sexually exploits or endangers children, in any form;
• incites or facilitates violence, terrorism or self-harm, or harasses, threatens or defames a person;
• deceives people at scale — spam, phishing, fraud, impersonation, undisclosed political or commercial manipulation, fake reviews;
• generates malware or provides operational uplift for attacks on systems, weapons capable of mass casualties, or critical infrastructure;
• makes automated decisions with legal or similarly significant effects on people (credit, employment, housing, insurance, law enforcement) without human review and the safeguards the law requires;
• collects or infers sensitive personal data about people without a lawful basis, or tracks people without their knowledge;
• presents model output as a human where the context makes that deceptive, or as professional medical, legal or financial advice without qualified review.

3. Protecting the platform

• Do not probe, scan, load-test without permission, reverse-engineer or attempt to bypass rate limits, budgets, guardrails, authentication or billing.
• Do not try to discover, target or interfere with the upstream providers and credentials behind a route; the service exposes the model vendor, not its infrastructure.
• Do not share one API key or Cursor service credential between several people, resell raw access to the endpoint as a bare proxy, or operate accounts to evade a suspension or a limit. Building products on top of the API is of course allowed.
• Do not send traffic designed to exhaust capacity (excessive parallelism, retry storms, deliberately oversized prompts) beyond what your key’s rate limit and plan allow.
• Do not upload content you have no right to process, including other people’s secrets or credentials.

4. Model vendor policies

Every request is fulfilled by a third-party model, and each vendor publishes its own usage policy (Anthropic, OpenAI, Google, DeepSeek and others). Their rules apply to requests that reach their models through ElevenRouter, including restrictions on using outputs to train competing models and on specific high-risk uses. A vendor may refuse a request or flag an account; repeated refusals for the same reason count as a violation of this policy.

5. High-risk applications

Model outputs can be wrong. If you deploy ElevenRouter in a context where an error could cause physical, financial or legal harm — medical, safety, legal, financial or infrastructure decisions — you must keep a qualified human in the loop, test for your use case, and disclose to affected people that an AI system is involved where the law requires it.

6. Enforcement

We may throttle, suspend or terminate keys or accounts that violate this policy, remove stored content, and preserve or disclose information when the law requires it. We consider the severity, whether the violation was deliberate, and whether it stopped once you were told. We notify you unless doing so is prohibited or would endanger others, and we lift a suspension once the cause is fixed. Unused purchased credits remain refundable under the Refund & Cancellation Policy except where used to cause harm.

7. Reporting abuse

To report content or activity that violates this policy, or a security issue, write to hello@elevenrouter.com with as much detail as you can (time, generation id or URL, what you observed). We acknowledge reports within two business days.