Public beta policy

Acceptable use policy

Effective August 29, 2026. This policy applies to every account, API key, request, application, and downstream user of API Many. Upstream provider policies also apply to requests routed to that provider; the more protective rule controls where requirements differ.

1. Illegal activity and harm

Do not use the service to plan, facilitate, automate, or conceal illegal activity or material harm. Prohibited uses include child sexual abuse or exploitation; non-consensual intimate content; credible threats or instructions intended to cause serious physical harm; trafficking; terrorism; unlawful weapons activity; and content that unlawfully discriminates against, harasses, or targets a person.

2. Fraud, deception, and abuse

Do not use the service for phishing, credential theft, scams, spam, fake reviews or engagement, deceptive impersonation, malicious social engineering, forged evidence, election interference, market manipulation, or evasion of sanctions, identity checks, or platform enforcement. Synthetic content must be disclosed when law, a platform, or the context requires it.

3. Malware and security misuse

Do not create, improve, deploy, or conceal malware; steal or expose credentials; exploit systems without authorization; disrupt services; or perform high-volume reconnaissance against targets you do not own or have permission to test. Security research involving API Many requires prior written permission. See the security page for reporting instructions.

4. Privacy, rights, and sensitive decisions

Do not submit personal, confidential, biometric, health, financial, educational, employment, or other regulated data unless you have a lawful basis, required notices and consents, and written confirmation that the eligible route meets your data requirements. Do not violate copyright, trademark, publicity, confidentiality, privacy, or other rights.

Do not use model output as the sole basis for decisions about medical care, legal rights, credit, insurance, employment, housing, education admission, law enforcement, immigration, or another high-impact area. A qualified human must review relevant information and remain responsible for the decision.

5. Platform and credential integrity

Do not share or expose an API key in client-side code; transfer accounts; create accounts or rotate keys to evade limits; interfere with metering; bypass safety systems, access controls, route restrictions, or rate limits; probe private provider information; scrape the service at abusive volume; or intentionally create workloads designed primarily to impose disproportionate cost.

You may not resell, sublicense, proxy, or provide shared access to API Many without a written agreement. A catalog entry is not permission to access or resell a model. Only a model returned by GET /v1/models is callable, and a paid route is enabled only after required route authorization is verified.

6. Enforcement

We may block a request, reduce limits, revoke a key, preserve relevant evidence, or suspend an account when reasonably necessary to investigate or stop a violation, comply with law or provider requirements, protect users or infrastructure, or address payment risk. Where appropriate, we consider context, intent, severity, recurrence, remediation, and legal obligations.

If you believe enforcement was a mistake, email [email protected] from the account email with the request ID and relevant context. Do not include an API key or unrelated sensitive content.

7. Reporting abuse

Report suspected abuse to [email protected]. Include URLs, API Many request identifiers, timestamps, and a factual description when available. Do not send illegal content itself unless we specifically request a lawful transmission method.

8. Changes

We may update this policy as the public beta, provider requirements, and law develop. Material changes will be posted with a new effective date.