Acceptable Use Policy

CheapWeights provides access to AI models, including models without built-in refusal behavior. What you generate is your responsibility; what's illegal stays illegal regardless of the tool.

You may not use the service to:

1. Break the law. Any content that is illegal in your jurisdiction or ours.
2. Harm people. Harassment, doxxing, stalking, threats, non-consensual intimate imagery, or sexual content involving anyone under 18 — in any form, fictional or otherwise.
3. Weapons and attacks. Instructions for weapons of mass destruction, explosives, or attacks on people or critical infrastructure.
4. Malware and intrusion. Generate, improve, or deploy malware, phishing kits, or exploits against systems you don't own or lack written authorization to test.
5. Fraud and deception at scale. Scams, disinformation campaigns, impersonation of real people, or fake identities presented as real.
6. Abuse the platform. Circumvent rate limits, share or resell keys, scrape the catalog, or probe other tenants.

Security research

Uncensored models are legitimate tools for red-teaming, safety evaluation, and refusal-mechanism research. Authorized testing of systems you own or have written permission to test is explicitly allowed.

Enforcement

Accounts are prepaid and rate-limited. We don't read prompts and we don't store them, so enforcement runs on account-level signals: request velocity, spend patterns, payment-fraud signals, reports. Violations draw a warning, throttle, suspension, or termination without refund, proportionate to severity. Illegal content means immediate termination and, where the law requires, reporting.

Reporting

abuse@cheapweights.ai — reviewed by a human.

Changes

Material changes are announced on the site 14 days before taking effect. Continued use after that means acceptance.