Usage Policy

Use models safely, respect others’ rights, and comply with platform and provider requirements.

Service provider · ThinkApiAI Inc

01Scope

This policy applies to requests submitted through ThinkAPI, generated content, and applications built with these capabilities. Follow applicable law, this policy, and the current usage policy of the model you call. Provide appropriate notices, access controls, and abuse safeguards for your authorized end users.

02Prohibited unlawful and harmful uses

Using the service to commit, facilitate, or conceal unlawful acts, including fraud, human trafficking, terrorism, or making or using weapons to harm people.
Creating, obtaining, or distributing child sexual exploitation material, grooming children, or otherwise endangering minors.
Violating privacy, intellectual property, or personal rights, including unlawfully obtaining personal data, non-consensual intimate imagery, impersonation, threats, or sustained targeted harassment.
Producing or distributing content intended to deceive, including phishing, spam, sham transactions, identity fraud, or manipulation of important information about public safety or election procedures.
Inciting violence, hatred, or persecution against people or groups, or encouraging suicide, self-harm, or other serious physical harm.

03System security and resource use

Do not access, scan, or attack systems without authorization, distribute malware, steal credentials, or disrupt ThinkAPI or third parties. Legitimate security research must stay within explicit authorization.
Do not bypass authentication, billing, balances, quotas, access controls, or content safeguards. Do not use bulk accounts, key rotation, or false source information to evade platform or provider restrictions.
Do not falsify payments or abuse referral rewards or promotional credits. Automated requests must respect applicable concurrency and rate limits and use reasonable retry controls.
Do not expose keys in public code, websites, or other public channels, or share accounts with unauthorized people. Keep access credentials in environments where access can be controlled.

04Applications and consequential decisions

When using models in products for others, accurately disclose AI involvement and provide appropriate user safeguards and feedback channels. Do not falsely present ThinkAPI or model outputs as officially certified, human expert conclusions, or guaranteed accurate results.

For uses that may significantly affect individuals, including health, law, finance, employment, and education, comply with additional provider requirements for high-risk uses and ensure necessary professional review, human oversight, and appeal mechanisms. Model outputs must not replace judgments legally reserved for qualified people.

Commercial integrations and end-user features remain subject to this agreement. Do not impersonate an official provider service or use resale, distribution, or proxy access to supply restricted models to ineligible parties.

05Enforcement and review

Depending on the nature, impact, and recurrence of conduct, we may require remediation, refuse requests, restrict rates, disable keys, or suspend or terminate an account. Measures needed to address ongoing harm or legal obligations may take effect immediately.

Payment does not exempt anyone from this policy. You may request review through support with necessary redacted records. Balances and refunds are handled under the Terms of Service and applicable law and are not automatically forfeited because of an unverified report.

06Model policies and reports

When calling Claude or GPT models, also follow the respective provider’s policies. A successful response does not establish that a use is permitted. Report abuse, security issues, or potentially infringing content through support, without including unnecessary sensitive information.