Acceptable use.

1. Allowed use

You may use the API from your own server or product for lawful text generation, transformation, classification, summarisation, coding, research, and other uses consistent with the Terms of Service, model documentation, and applicable law. You remain responsible for the applications, prompts, outputs, users, and decisions connected to your account.

2. Prohibited use

You must not use AddBrains to:

• break the law, infringe rights, deceive people, facilitate violence, or evade sanctions;
• create or distribute malware, credential theft, phishing, ransomware, or instructions intended to enable cyber abuse;
• exploit, sexualise, or endanger children, or generate non-consensual intimate material;
• target a person or group with harassment, threats, discrimination, or unlawful surveillance;
• impersonate a person or institution, generate fraud, or conceal synthetic material where disclosure is required;
• provide medical, legal, financial, employment, housing, education, insurance, or public-service decisions without appropriate qualified human oversight;
• train, benchmark, or extract a competing model or dataset in a way that violates the Terms, a model licence, or third-party rights; or
• send high-volume automated traffic designed to evade controls, overload the service, or create avoidable provider cost.

3. Content and rights

Do not submit material unless you have the rights and lawful basis needed for AddBrains and the production upstream to process it. Do not submit secrets, payment-card data, authentication tokens, or unnecessary personal data. If your workflow handles personal data, you must use appropriate safeguards and provide any notice required to the people concerned.

Model output can be inaccurate, biased, unsafe, or similar to output provided to others. Review it before relying on it or sharing it. AddBrains does not make professional decisions for you.

4. High-impact decisions

Do not use model output as the sole basis for decisions that determine a person’s access to employment, housing, credit, insurance, education, essential public services, medical treatment, or legal rights. Qualified humans must independently review high-impact outcomes, and you remain responsible for sector-specific requirements and records.

5. Accounts and credentials

• Keep API keys server-side. Do not sell, share, publish, or transfer them.
• Do not create accounts for banned users, evade suspension, or misrepresent customer classification.
• Revoke a key you believe is exposed and contact support without sending the key.
• Do not resell raw AddBrains access unless a separate written agreement expressly permits it.
• Do not use internal, promotional, or test grants as a way to avoid payment or usage limits.

6. Technical abuse

Do not bypass authentication, rate limits, spend caps, model restrictions, safety controls, margin guards, request-size limits, concurrency limits, or launch flags. Do not probe for secrets, attack other customers, interfere with metering or the ledger, scrape non-public surfaces, reverse engineer protected systems, or design requests to create an uncertain provider or payment outcome.

7. Model and upstream rules

Valid published restrictions of the selected model and Cheaper route may apply when clearly incorporated into the service documentation. AddBrains does not publicly disclose confidential provider material. If a restriction conflicts with mandatory law, mandatory law controls. OpenRouter is a reference-rate source only and is not a request route.

8. Enforcement and review

AddBrains may rate-limit, block a request, revoke a credential, restrict a model, suspend an account, preserve necessary evidence, or report conduct where reasonably necessary to protect people, providers, the service, or the law. Action should be proportionate to the risk. Where lawful and safe, we will provide a reason and a route to request human review. Mandatory consumer rights are not excluded.

9. Report a concern

Email [email protected] with a request identifier, model alias, date, and a description that does not include an API key or unnecessary personal data. If you believe a restriction is mistaken, ask for review through the same address. For privacy concerns, use [email protected] and see the Support and Privacy Requests page.