Acceptable use policy

The models on this service have had their refusal behaviour removed. They will answer requests that other providers decline. That makes you, the customer, responsible for how you use them.

You may not use the service to

• Generate sexual content involving minors, or any content that sexualises minors.
• Create non-consensual intimate imagery or text about real people.
• Plan, facilitate or carry out violence, terrorism, or attacks on critical infrastructure.
• Seek meaningful help building biological, chemical, nuclear or radiological weapons.
• Write or deploy malware, or attack systems you are not authorised to test.
• Harass, threaten, dox, or impersonate real people, or run fraud and scams.
• Break any law that applies to you or to the people your application serves.

Your responsibilities

• You are responsible for all output generated with your API keys, and for filtering it before showing it to your own users where that is appropriate.
• Keep your keys secret. Usage on a key is billed to its account.
• Don't resell raw access to people who haven't agreed to this policy.

Enforcement

One platform rule applies to every account and can't be turned off: sexual content involving minors. A request or reply that matches is blocked, and we keep a record that it happened, without its text. Nothing else is blocked automatically, so filtering for your own users, including people who may be at risk of self-harm, is up to you. Automated checks make mistakes in both directions, so this doesn't change your responsibilities above.

We may suspend or close accounts that break this policy, including accounts that repeatedly hit the platform rule, and we cooperate with lawful requests from authorities.