The short version: FaKnows is entertainment, not a professional opinion. It's generated by AI and can be wrong. Don't upload a chat to harass or expose someone, and don't treat the verdict as real advice about your relationship, a legal matter, or anything else that actually matters.
FaKnows reads a chat export you upload and generates an AI-written "verdict" — a balance score and a short, opinionated take on the dynamic between the people in it, backed by quotes pulled from your own chat. It's styled as entertainment. It is not therapy, not a clinical assessment, not legal advice, and not a diagnosis of anyone.
Verdicts are written by Anthropic's Claude or OpenAI's ChatGPT (see our Privacy policy for which, and when). Like any AI output, a verdict can be wrong, unfair, or miss context the model never had. We put real effort into keeping scores and counts accurate — balance scores and message counts are computed in code from your chat, not guessed by the model — but the written verdict itself is a generated opinion, not a fact.
We rate-limit requests per feature to keep the service usable for everyone. Automating requests, scraping, or attempting to bypass rate limits or the ₹99 unlock is not allowed.
FaKnows doesn't require an account. The service is provided "as is" — we don't guarantee it will be available every moment, free of bugs, or that a given verdict will be accurate, fair, or useful. See our refund policy for what happens if a paid unlock doesn't work as expected.
The full report is a one-time ₹99 unlock processed by Razorpay. See our refund policy.
We may update these terms as the product changes. The "last updated" date above will always reflect the latest version.
Questions about these terms: [email protected].