About Clair

Clair is built and operated by Serin Technologies, LLC, a Delaware limited liability company. We write and maintain the code that turns each source's page into fields ourselves, and we answer support email ourselves.

Reach us at contact@serinlabs.com or on X at @clair_api.

• Your call names a page: a search, a product, a company, a post. Clair fetches that page from the source as a signed-out visitor would see it.
• Most of these sites refuse requests from a data center, so pages are fetched over residential proxies, and rendered in a real browser only when the page needs it. How a page is fetched does not change the price you pay.
• Our own parser reads the page and returns the documented fields as JSON, in the same shape however the page was fetched. The parsers are the part of Clair we build; the response examples on each API page are recorded from real calls.
• A fetched page is kept for at most ten minutes, so a repeat of the same request inside that window returns the same page. There is no long-term store or history: Clair holds no database of the sources' data.
• A request is billed when it is answered. A request that fails on our side (5xx) is refunded to your balance.

• Log in. Clair never signs in to a source, so anything a site shows only to members, such as Glassdoor's full review history or posts Instagram hides from signed-out visitors, is not returned.
• Build an API where the source already gives developers the same data for free through its own API. We write a parser only where an official API is closed, retired, restricted, or priced far above the page itself, and each API page says which, with a link to the source's own documentation.
• Collect sensitive categories of personal data about individuals, or let results be used to message people without the consent the law requires. The terms of service set these out.

The full rules for using the results are in the terms of service and the privacy policy.

• Statements about a source's own API, such as whether Glassdoor or Indeed still offer one, link to the source's documentation and show the date we last read it.
• Comparison pages cite the competitor's pricing and documentation pages with the date they were checked, and each lists the cases where the competitor is the better choice.
• Clair's prices, on the pricing page, the API pages, and the comparisons, are computed from the same billing code that charges your balance. None of them is typed in by hand, so they cannot drift from what you pay.
• Latency and coverage figures come from benchmark runs we record with their date and query count. If a figure has not been measured, the page does not show one.

If you find something wrong on this site, email contact@serinlabs.com and we will correct it.

Every hour, a scheduled check calls each important endpoint with a known-good request and holds the response to minimum field counts, so a broken parser shows up before customers report it. The status page reports 30-day uptime and response times for every API from those checks and from customer requests.