Acceptable Use Policy

Don't use NeverMetered to create, request or distribute:

• Child sexual abuse material, or any sexual content involving minors. We have zero tolerance: accounts are closed immediately and we report to the relevant authorities.
• Content or activity that is illegal where you or your users are, or that facilitates it, such as fraud, trafficking, or buying and selling illegal goods.
• Harassment, threats, bullying, or content that promotes violence or hatred against people based on protected characteristics.
• Content that promotes or supports terrorism or violent extremism.
• Non-consensual intimate imagery or sexual content about real people without their consent.
• Instructions intended to cause serious harm, such as weapons capable of mass casualties.

• Writing, improving or deploying malware, ransomware or phishing kits.
• Exploiting, scanning or attacking systems you don't own or aren't authorised to test. Authorised security research, CTFs and defensive work are fine.
• Credential stuffing, account takeover, or cracking passwords you have no right to.

• Spam, including bulk unsolicited messages, comment spam or SEO content farms.
• Scams, fraud, fake reviews, and impersonating real people or organisations to deceive.
• Election or civic disinformation, or content designed to deceive people about how, when or where to vote.
• Passing off AI-generated content as human-written where the law or platform rules require disclosure.

• Generating or distributing content that infringes copyright, trademarks or other rights.
• Using a model in a way its licence prohibits. Each model we serve is published under its own licence; you must follow it (see the Terms).

Don't use outputs to make automated decisions with legal or similarly significant effects on people (medical diagnosis or treatment, legal advice, credit, insurance, housing, employment or education decisions) without qualified human review. Models make mistakes.

NeverMetered is shared capacity, and plans are priced on the promise that everyone plays fair. Don't:

• Create multiple accounts to get extra free requests or get around limits, suspensions or bans.
• Share, publish or resell API keys, for example by committing them to public repositories or embedding them in apps others can extract them from.
• Resell raw access to the API (for example as a proxy or “API reseller”) without our written permission. Building your own products, agents and apps on top of NeverMetered is fine and encouraged.
• Try to access other users' accounts, data, responses or cache, or to infer their content.
• Deliberately flood the service with traffic beyond your plan, or run load tests or benchmarks designed to degrade it for others, without agreeing it with us first.
• Probe, scan or attack the service, bypass its authentication, rate limits or abuse protections, or interfere with its operation.

Normal heavy use, like coding agents, batch jobs and long contexts within your plan's limits, is exactly what the service is for.

If you put NeverMetered behind your own app or service, you're responsible for how your users use it. Make sure your own terms prohibit the uses above, and act on abuse you become aware of. Give background jobs and end-user traffic their own API keys so you can revoke one without affecting the rest.

If we believe this policy has been broken, we may, depending on severity:

• contact you and ask you to stop;
• throttle traffic or lower your limits;
• revoke API keys or block free requests;
• suspend or close the account, and related accounts used to get around enforcement; and
• report illegal activity to law enforcement or other authorities.

Because we don't store prompt content, enforcement relies on usage patterns, account signals and reports. If you think we've made a mistake, reply to our email and we'll review it.

Report abuse, leaked keys or security issues to legal@nevermetered.com. Include as much detail as you can (times, request ids from response headers, and what happened). Please don't test vulnerabilities in ways that affect other users.