Terms of use
Short, because the service is free and the rules are few.
You may
- Use the API for anything, including commercial products, with no upgrade or permission.
- Store and process the data you retrieve.
- Call it from a server, a browser, a phone or an autonomous agent.
- Do all of the above without telling us or registering.
You should
- Link back. When you show an article to a person, link to the original
url. It is the publisher's work, and it is the only thing we ask. - Cache. Data changes hourly. Repeating the same query every second helps nobody.
- Stay within one identity. The 20 requests per second limit is per IP. Spreading a bulk export across many addresses to get around it is the one behaviour that would force us to add keys.
You may not
- Present article content as your own work, or strip attribution.
- Use the service to build a product whose purpose is to deceive people about what news sources reported.
What is guaranteed
Nothing. There is no SLA, no uptime commitment, no support obligation and no warranty of accuracy. The service has been reliable and we intend to keep it that way, but it is provided as-is and free of charge. If your business needs a guarantee, buy one from a vendor who sells them — several do.
Content ownership
Article text belongs to the publishers who wrote it. We extract it from a public crawl and index it for search under fair-use-style access; we claim no ownership and grant no licence to it. Publishers can have their content removed at any time — see for publishers.
Changes
Free keyless search is permanent. Any other change is announced on the changelog before it takes effect.
