1. Purpose
TokensMind provides API access to AI models and related routing services. Users must not use TokensMind to create, facilitate, automate, scale, conceal, or distribute harmful, illegal, deceptive, abusive, or rights-infringing activity.2. Prohibited activity
You must not use the Services for:- Fraud, scams, phishing, impersonation, fake customer support, credential theft, payment fraud, or social engineering.
- Malware, ransomware, spyware, exploit code, credential stuffing, unauthorized access, botnet activity, vulnerability exploitation, or evasion of security controls.
- Spam, bulk unsolicited messaging, fake reviews, engagement manipulation, account farming, platform manipulation, or automated harassment.
- Child sexual abuse material, sexual exploitation of minors, grooming, or content sexualizing minors.
- Non-consensual sexual content, non-consensual intimate imagery, abusive deepfakes, or sexual extortion.
- Adult sexual services or explicit sexual content where illegal or where it violates supplier policies.
- Non-consensual biometric identification, face recognition, voice cloning, impersonation, or misuse of another person’s likeness.
- Copyright infringement, trademark infringement, piracy, unauthorized generation of protected characters or brands, or violation of publicity, privacy, or personality rights.
- Terrorism, violent extremism, recruitment, propaganda, instructions for serious violence, weapon construction, or evasion of law enforcement.
- Hate, harassment, threats, doxxing, intimidation, or unlawful discrimination against protected groups.
- Illegal gambling, illegal drugs, illegal weapons, sanctions evasion, money laundering, or regulated services without authorization.
- Medical, legal, financial, employment, housing, credit, insurance, or other high-impact decisions without required human review and legal compliance.
- Circumventing model provider safety systems, account limits, geographic restrictions, sanctions controls, rate limits, or monitoring systems.
- Reselling API access, sharing accounts, pooling supplier accounts, or providing unauthorized downstream access in violation of supplier terms.
3. Content categories covered
This Policy applies to text, code, images, audio, video, embeddings, files, prompts, outputs, automated agents, workflow automation, and any other content or activity processed through the Services.4. Safety controls
TokensMind may use one or more of the following controls:- Account, email, phone, organization, device, IP, country, BIN, card, payment, and behavioral risk checks.
- API key monitoring and anomaly detection.
- Rate limits, concurrency limits, spend limits, model limits, and velocity controls.
- Keyword, pattern, endpoint, model, token-volume, and usage-spike detection.
- Upstream model provider safety filters, moderation systems, and abuse reports.
- Manual review of support tickets, user reports, dispute materials, metadata, and provider alerts.
- Temporary holds, step-up verification, order review, API key revocation, account suspension, account termination, or blocking of requests.
- Evidence retention for fraud, abuse, chargeback, legal, regulatory, or law enforcement purposes.
5. Prompt and response handling for safety
TokensMind does not persistently store prompt or response content by default. Abuse investigations primarily rely on metadata, usage records, payment records, API key records, IP/device information, provider safety signals, user reports, and materials voluntarily submitted to support. Where legally permitted and necessary, we may temporarily preserve or review limited relevant materials for serious abuse, fraud, security incidents, payment disputes, legal requests, or enforcement of this Policy.6. User obligations
You must:- Use the Services only for lawful and policy-compliant purposes.
- Implement reasonable safeguards in your own products and workflows.
- Monitor downstream users if you provide access through your application.
- Keep API keys secure and rotate compromised keys promptly.
- Respect supplier terms, safety policies, and rate limits.
- Respond promptly to TokensMind inquiries about suspected misuse.