Last updated 1 October 2026.
oassis is an API that fetches web pages on behalf of whoever calls it. This page says exactly what we keep while doing that. It describes the system we actually run, and it changes when the system does.
| What | Why | Kept |
|---|---|---|
| Account id, balance, API key hash | To charge the right balance | Until the account is deleted |
| One row per billable call: time, amount, route, reference | So a charge can be explained and disputed | Until the account is deleted |
| Browser sessions and jobs: when opened, how much charged | Billing and support | Until the account is deleted |
| Feedback you choose to send | To fix what is broken | Until acted on |
| Rendered pages, cached | So a repeat read is cheaper for you | Short-lived, then discarded |
| A one-way fingerprint of the caller's IP address | To count free calls without accounts | Rolling daily counters |
| A short request log: time, route, status, how long it took, your IP, your user-agent, and a truncated copy of the request and the answer | So a call can be explained when something goes wrong | The most recent 2,000 requests, and never more than 14 days |
Calls made with no key and no account are counted against a one-way hash of the IP address, truncated. We do this so a free allowance can exist at all without asking anyone to sign up. The address itself is not stored in that counter, and the hash cannot be turned back into it.
We keep a short operational log of recent requests so that a call can be explained when something goes wrong. It is bounded in both directions: at most 2,000 requests are kept at any time, nothing older than 14 days survives, and the bodies are stored truncated — enough to see what was asked for, not a copy of the answer. Large responses are recorded as a size and never read. It is read by us, to run the service, and by nobody else.
We request the url you give us and return the result to you. The content is held only as long as it takes to answer, plus a short cache so that asking twice costs you less. We do not keep a copy of scraped content in our database, and we do not build a corpus out of what our users read.
We do not sell data, we do not share it for advertising, and we do not pass it to anyone not listed above.
oassis is a developer tool and is not directed at children.
If this policy changes, the date at the top changes with it.
Write to support@oassis.dev. A person reads it.