Sitemap to IndexNow

A site on www.example.com deploys a new build. The old sitemap had 7 URLs; the new one has 8. Compared URL by URL:

• Added (2): /blog/new-post and /features are only in the new sitemap.
• Removed (1): /blog/old-post is only in the old one. It should now return 404 or 410, or redirect, and submitting it tells search engines to drop or update it.
• Changed (1): /pricing is in both, but its lastmod moved from 2026-09-01 to 2026-09-30.
• Unchanged (5): the other pages have the same lastmod in both, so they are not sent again.

That is 2 + 1 + 1 = 4 URLs, all on one host, so ⌈4 ÷ 10,000⌉ = 1 request, with a 305-byte body:

POST https://api.indexnow.org/indexnow
Content-Type: application/json; charset=utf-8

{"host":"www.example.com","key":"3f9c1a7e5b2d48f0a6c4e8b1d7f2a905",
 "keyLocation":"https://www.example.com/3f9c1a7e5b2d48f0a6c4e8b1d7f2a905.txt",
 "urlList":["https://www.example.com/blog/new-post","https://www.example.com/features",
            "https://www.example.com/pricing","https://www.example.com/blog/old-post"]}

With Changed since instead, time zones matter: a lastmod of 2026-09-19T23:30:00-02:00 is 2026-09-20 01:30 UTC, so it passes a cutoff of 2026-09-20 00:00 UTC even though the date written in it is the 19th.

The common mistake: sending everything in one request with "host": "www.example.com" when some URLs are on example.com. IndexNow treats www and non-www as different hosts, so the request fails with 422. The “www + non-www list” preset shows the fix: 4 www.example.com URLs and 2 example.com URLs become 2 batches, and example.com needs its own key file. The other habit to avoid is resubmitting the whole sitemap on every deploy; send only what changed.

Google introduced Sitemaps in 2005 so site owners could list pages its crawler might miss. In 2006 Yahoo and Microsoft adopted the same format, and the three published it as a shared protocol at sitemaps.org. A sitemap is a pull model: it lists every URL you want indexed, and crawlers read it on their own schedule, so a change can take days to be noticed.

IndexNow, launched in 2021 by Microsoft Bing and Yandex, is the push model. The site sends a short list of URLs that just changed, with a key that proves it controls the host, and the participating engines can crawl them straight away. Submissions are shared, so one request reaches Bing, Yandex, Seznam.cz, Naver, Yep, Amazon and the Internet Archive. The two work together: the sitemap is the complete list, IndexNow announces the changes, and this tool turns the first into the second.

This tool reads sitemaps (including gzipped files and sitemap indexes), plain URL lists and robots.txt files, and turns them into IndexNow submissions. It can submit every URL, only those whose lastmod is on or after a date (time zones converted to UTC), or the added, removed and changed URLs between two versions of a sitemap. URLs are checked, normalised with the browser’s URL standard, filtered by path and de-duplicated, then grouped by host and split into batches of up to 10,000.

The output is the exact JSON body for each batch and ready-to-run cURL, PowerShell, Python, PHP and Node.js code that prints each response and stops on an error with the reason. It also generates and validates your key and saves the key file. Everything runs in your browser; nothing you paste or drop is uploaded, and no request is sent to IndexNow from this page. For the full setup, including automation and troubleshooting, read How to Set Up IndexNow.

Related tools: XML Sitemap Generator, robots.txt Generator, and cURL Converter.