Privacy policy

Template: review with counsel and replace bracketed fields before launch.

[Company legal name] ("we") operates this API. This policy explains what we process and keep.

Prompts and outputs

We do not log, store, or review prompts, completions, or tool data.
We do not use your data to train models, and we do not sell it.
While a request runs, its content exists in server memory and GPU memory. Reusable prompt prefixes stay in a cache that only your organization can hit, until evicted (usually minutes, at most until the next engine restart).
Requests routed to us by partners (for example NanoGPT) follow the same rules.

Account data: email, name, password hash, organization membership, sessions (IP address and user agent of the session).
Billing records: request ids, timestamps, model deployment, token counts, charges, credit purchases and refunds. Kept for [N] years for tax and accounting.
Payment details are handled by our processors ([Stripe], [NOWPayments]). We receive payment status and amounts, not card numbers.
Operational metrics without request content (latency, error counts).

Infrastructure: [GPU host], [control-plane host], [backup storage], in [countries]. Email: [provider].

Ask for a copy or deletion of your account data at support@infrenis.com. Billing records required by law are kept for the legal period.

We respond to valid legal process. Because we do not store prompts or outputs, we cannot produce them.

Last updated: [date]. Jurisdiction: [Jurisdiction].