What is data retention?
Data retention is how long an API keeps what it produced for you: scraped pages, extracted results, logs and browser session recordings. It is often set by plan, from a few days to a few months, and some terms also say what the vendor or you may do with the data afterwards.
Data retention across the APIs we track
| API | What the vendor publishes | Read |
|---|---|---|
| Anchor Browser | The docs plan table says 45-day retention for Free and Starter, while the homepage says 30-day retention for Team. | Jul 26, 2026 |
| Brave Search API | All results must be deleted when the agreement ends. | Jul 26, 2026 |
| Browse AI | Free, Personal, and Professional retain data for 90 days. | Jul 26, 2026 |
| BrowserCloud | Data retention is 7 days on Free and 30 days on Hobby, Startup, and Business. | Jul 26, 2026 |
| Kernel | Hobbyist retains browser replays for 7 days and Start-Up for 30 days. | Jul 26, 2026 |
| Nimble | Data retention is 7 days on Startup, 30 days on Scale, and 90 days on Professional. | Jul 26, 2026 |
| Perplexity Search API | The Search addendum says Perplexity may retain, copy, distribute, and use Search Data for lawful business purposes, including product improvement. | Jul 26, 2026 |
| Perplexity Search API | The generic API FAQ says API query data is not retained, which conflicts with the later Search-specific addendum. | Jul 26, 2026 |
| SearchApi.io | Zero retention is an enterprise-only option; standard cached search HTML and JSON follow SearchApi's retention schedule. | Jul 26, 2026 |
| SOAX Web Data API | Public Web Data API concurrency, data-retention, and over-quota behavior were not published on the loaded pages. | Jul 26, 2026 |
| Steel.dev | Launch allows 10 concurrent sessions, 15-minute maximum sessions, 3 seats, 60 requests per minute, and 7-day retention. | Jul 26, 2026 |
| Steel.dev | Scale allows 100 concurrent sessions, 1-hour maximum sessions, unlimited seats, 600 requests per minute, and 14-day retention. | Jul 26, 2026 |
| Tavily | Tavily advertises zero data retention. | Jul 26, 2026 |
| Webscraper.io | Data retention is 30 days on Project, 30 days on Professional and 60 days on Scale. | Oct 11, 2026 |
| You.com API | Zero data retention is available as a data-control option rather than being stated as the universal default. | Jul 26, 2026 |
Only vendors that publish the term are listed; a vendor missing here has not published it, which is not the same as having none.
In more detail
The vendor examples below were read on the dates given in the text; the table above holds each vendor's current figure.
Why it matters
If you read results from the vendor’s storage rather than saving them as they arrive, the retention period is your deadline. Short retention is fine for a pipeline that writes every result to its own database; it is a risk for a team that re-downloads past runs or needs records for an audit. Terms on what the vendor may do with your queries and results matter when the data is confidential, and terms on what you may do with the results matter when you plan to store or republish them.
The table quotes each vendor’s published terms, with the day each was read.