What Is IndexNow? How It Works, Who Uses It, and What It Can't Do

IndexNow is an open protocol that lets a website tell search engines a URL has been added, updated or deleted, instead of waiting for a crawler to notice. You host a key file on your domain to show you control it, send the changed URLs to one endpoint, and every participating engine receives them: Microsoft Bing, Naver, Seznam.cz, Yandex and Yep are the search engines listed on indexnow.org. Google doesn't take part, and an accepted submission means the engine has your URL, not that it crawled or indexed the page.

What IndexNow is

Search engines find changes by recrawling URLs they already know, and they don't revisit every URL often. The IndexNow site says that without a notification, a change can take days to weeks to be discovered. IndexNow reverses the direction: when something changes, your site pings the engines with the URL, and they decide whether and when to crawl it. A sitemap is pulled by the engine on its schedule; an IndexNow notification is pushed by you at the moment of the change.

Microsoft Bing and Yandex created the protocol, and Bing announced it on October 18, 2021 (Bing Webmaster Blog). That's why it's often called Microsoft IndexNow, and Bing still builds much of the tooling around it, including the official WordPress plugin and the IndexNow reports in Bing Webmaster Tools. The protocol itself is open: indexnow.org publishes it under a Creative Commons Attribution-ShareAlike license, and Bing's Why IndexNow page describes it as a free, open-source protocol. The protocol asks for no account with any engine: the key file is the proof of control.

It is widely used. In December 2024, Bing reported more than 3.5 billion URLs submitted per day through IndexNow, and put the protocol's share at 18% of all new URLs clicked in web search results. The same post named GoDaddy, the Internet Archive and Condé Nast among recent adopters. As of October 2026, the Why IndexNow page shows sites and platforms such as LinkedIn, eBay, Etsy, GitHub, Wix, Yoast and Cloudflare among those adopting it.

How the IndexNow protocol works

The IndexNow protocol has three parts: a key, a key file that proves you control the host, and the submission itself. The full rules are in the IndexNow documentation.

1. Generate a key

The key is a string of 8 to 128 characters made of lowercase and uppercase letters, digits and dashes. Any random string within those rules works, for example 32 hex characters or a UUID. One key per host is enough. If you ever need a new key, the IndexNow FAQ says to replace the key file and start submitting with the new key: engines re-verify when they see it.

2. Host the key file

Put a UTF-8 text file named after the key, containing the key, at the root of the host: https://www.example.com/b6e3f0c94d2a4c1e9f7a8d5b3c2e1f00.txt. That's the recommended option. You can also host the file elsewhere on the same host and pass its URL as keyLocation, but the location then limits what the key covers: a key file under /catalog/ vouches for URLs under /catalog/ and not for URLs under /help/.

The FAQ treats each subdomain as a separate host, so blog.example.com needs its own key file. Keep the file online for as long as you submit: engines fetch it to check ownership, and Yandex's IndexNow help says it checks the key on every request. IndexNow key covers generating and hosting it on hosts and frameworks where the root is awkward to write to.

3. Submit the URLs

For one URL, send a GET request with the URL-encoded address and your key:

curl "https://api.indexnow.org/indexnow?url=https%3A%2F%2Fwww.example.com%2Fproducts%2Fblue-mug&key=b6e3f0c94d2a4c1e9f7a8d5b3c2e1f00"

For several, POST a JSON body. One request can carry up to 10,000 URLs, http and https can be mixed, and every URL must belong to the host you name:

curl -X POST "https://api.indexnow.org/indexnow" \
  -H "Content-Type: application/json; charset=utf-8" \
  -d '{
    "host": "www.example.com",
    "key": "b6e3f0c94d2a4c1e9f7a8d5b3c2e1f00",
    "keyLocation": "https://www.example.com/b6e3f0c94d2a4c1e9f7a8d5b3c2e1f00.txt",
    "urlList": [
      "https://www.example.com/products/blue-mug",
      "https://www.example.com/blog/new-post"
    ]
  }'

keyLocation is only required when the key file isn't at the root. The IndexNow API guide has the full request format and code examples.

4. The engine verifies the key and shares the URLs

The engine that receives the request fetches your key file. According to the FAQ, a first request can come back 202 Accepted while that check is pending, and later ones typically return 200 OK. Participating engines agree to share submitted URLs with each other, so a submission to api.indexnow.org or to any one engine reaches all of them. You submit once, not once per engine. Each engine then decides on its own whether to crawl and index the URL.

IndexNow response codes

None of these codes reports a crawl. An IndexNow 202 or 200 response explains what the success codes promise, and IndexNow 403, 422 and 429 errors has a fix for each failure.

Which search engines use IndexNow

As of October 2026, the indexnow.org home page lists five participating search engines: Microsoft Bing, Naver, Seznam.cz, Yandex and Yep. The FAQ's list of endpoints adds Amazon. Bing announced in May 2025 that Amazon planned to begin adopting IndexNow in mid-June of that year. The protocol's machine-readable list of participants also includes the Internet Archive, one of the adopters Bing named in December 2024.

Every endpoint takes the same request at the path /indexnow. The FAQ asks you to send each submission to only one of them, since the engines share it, so the shared endpoint is the simple default.

Each engine has its own webmaster tools for the rest of the job: Bing IndexNow covers the Bing Webmaster Tools side, and Yandex Webmaster, Naver Search Advisor and Seznam Webmaster cover the others.

Google doesn't use IndexNow

Google isn't on any of the indexnow.org lists of participants, so an IndexNow submission sends nothing to Google. In November 2021, Google said it would be testing the protocol's potential benefits, as Search Engine Journal reported at the time. As of October 2026, Google still isn't listed as a participant. Does Google use IndexNow follows that history and what to use for Google instead.

For Google, the channels are the ones it documents in Ask Google to recrawl your URLs: a sitemap for many URLs, and Request indexing in the URL Inspection tool for a few, subject to a quota. Google says a crawl request doesn't guarantee inclusion in search results, and that requesting the same URL repeatedly won't get it crawled any faster. Request indexing explains the button and its limit, and how to get your website to show up on Google covers the whole process.

Google's Indexing API is not a Google version of IndexNow. Its quickstart says it can only be used for pages with JobPosting or BroadcastEvent embedded in a VideoObject, and that every submission goes through spam detection. Using it for other pages is outside Google's documented use. More in Google Indexing API.

What IndexNow can't do

IndexNow is a notification. Getting a page into search results takes five separate stages, and IndexNow performs exactly one of them:

1. Discovery: the engine learns that a URL exists, from links, sitemaps or a notification.
2. Notification: you tell the engine that a URL was added, changed or removed. This is the step IndexNow does.
3. Crawling: the engine fetches the page, on its own schedule, or not at all.
4. Indexing: the engine decides to keep the page and can show it in results.
5. Ranking: the engine decides where the page shows, and for which queries.

The documentation says a 200 only indicates that the engine received your URL. The FAQ says submitting doesn't guarantee immediate indexing, that an engine may choose not to index a URL based on its algorithms, and names low-quality, thin and duplicate content among the reasons a URL isn't indexed. Yandex's help is blunter: the method doesn't guarantee that the pages you send will be indexed. In practice:

• It doesn't force a crawl. The FAQ says engines decide based on crawl quota, scheduling and quality signals, and that every URL you submit counts toward your site's crawl quota. Bing's getting started page adds that engines generally prioritize submitted URLs, but the engine still chooses.
• It doesn't fix a page. If a page has a quality or technical problem, a notification brings the crawler to the problem sooner. Whether the page is indexed and where it ranks are decided after the crawl, on what the engine finds there.
• It does nothing for Google. See the section above.
• It only works on hosts you control. The key file has to be on the host, and URLs that don't belong to the host in the request return 422. IndexNow can't be used for backlinks or third-party pages; backlink indexers covers that question.
• It doesn't replace sitemaps. The FAQ recommends both: IndexNow for recent changes, and a sitemap with accurate lastmod values for the full inventory, deeper pages and pages that rarely change. IndexNow vs sitemap compares the two, and XML sitemaps covers the sitemap side.
• It isn't a backfill. The FAQ says IndexNow isn't intended for submitting older changes retroactively. After a recent site-wide change, submitting all URLs is acceptable.

Whether it's worth setting up depends on how often your content changes and how much of your traffic comes from the participating engines. Is IndexNow worth it weighs that honestly.

What to submit, and how often

Submit a URL when it's added, when its content changes in a way that matters, and when it's removed. The FAQ's examples include new products or posts, changed descriptions or metadata, price or availability changes, and new reviews or comments. It also asks for URLs that now redirect (301, 302 or meta refresh) or return 404 or 410, so engines can update or drop them. Bing's guidance on when to notify adds detail per type of site: a shop should notify as soon as a product becomes unavailable, and review counts should be notified in tiers rather than on every new review.

What to leave out:

• Cosmetic edits. The FAQ asks automated scripts to detect real content changes, and not to submit the same URL many times a day without a meaningful change. It suggests waiting at least 5 minutes before resubmitting a URL in one answer and at least 10 minutes in another; either way, one submission per real change is the rule.
• URLs on other hosts. One request covers one host. Split submissions per host and subdomain, each with its own key file.

No daily cap is published. The FAQ says exact rate limits aren't disclosed and each engine sets its own thresholds per site. Past them you get a 429, and the advice is to pause, retry after the time in the Retry-After header, and send smaller, less frequent batches. The only fixed number is per request: 10,000 URLs.

How to use IndexNow on your site

How to use IndexNow depends on what runs your site. Bing's getting started page sorts the options into built-in support, plugins, CDNs and your own integration, and suggests checking Bing Webmaster Tools first, since your platform may already be sending updates.

Platforms with built-in support

As of October 2026, Bing lists Wix Studio, Duda, GoDaddy's website builder, XenForo, Milestone and Shopify as supporting IndexNow out of the box, with nothing to set up. Check what your platform already sends before adding an app or a second integration on top: IndexNow on Shopify covers Shopify stores, and IndexNow on Wix, Webflow, Squarespace and Framer covers the site builders.

WordPress plugins

The IndexNow plugin for WordPress is built by the Bing Webmaster team and had more than 100,000 active installations as of October 2026. It submits pages in the background when they're created, updated or deleted, skips URLs with a noindex directive, and sends everything to api.indexnow.org. Bing also lists Yoast, Rank Math, AIOSEO and SEOPress as options. Turn IndexNow on in one of them, not two, or every change goes out twice. IndexNow for WordPress compares the plugins.

Cloudflare Crawler Hints

If your site is behind Cloudflare, Crawler Hints sends IndexNow notifications for you based on cache signals: a cache MISS suggests the content may have changed. Cloudflare documents it as available on every plan, and it applies to the whole site once enabled. It infers changes from cache behavior rather than from your CMS, so it knows less about what actually changed than your own code would. Cloudflare IndexNow covers when that's enough.

Your own code

On a custom stack the integration is small: serve the key file, then POST the URLs that changed after each publish, deploy or rebuild. The hard part is knowing which URLs changed. Good sources are your CMS's publish events, the difference between the old and new sitemap after a build, and the paths your framework revalidated. Guides by stack: IndexNow in Next.js, IndexNow for static sites (Astro, Hugo, Jekyll, Eleventy), and IndexNow from a GitHub Action or deploy hook. To send notifications from a coding agent, see IndexNow MCP.

Check what Bing received

Bing Webmaster Tools has an IndexNow Insights view. When Bing introduced it in March 2024, it showed counts of URLs submitted, crawled and indexed, the latest 1,000 submitted URLs with their index status, first indexed date and whether they were submitted late, where submissions came from (such as Cloudflare, WordPress or manual submission), and errors grouped by cause. It is the closest you get to seeing what happened after a submission, for Bing. The wider picture, from Bing Webmaster Tools to sitemaps, is in our guide to Bing indexing.

How IndexCyborg uses IndexNow

IndexCyborg generates an IndexNow key for each website, and you host it at /<key>.txt on your domain. Nothing is sent until two checks pass: domain ownership, with a DNS TXT record or a hosted HTML file, and the key file, which has to return 200 with the key as its content. After that, each sitemap rescan finds new pages and pages whose lastmod changed, and IndexCyborg notifies them to api.indexnow.org automatically, up to 1,000 URLs per call. It skips URLs it notified recently, retries later after a 429 or a server error, and records every submission with the engine's response code; a failed submission shows that code in the activity log. The REST API (POST /v1/sites/{id}/notify) and the MCP tool notify_search_engines send URLs on demand, behind the same checks.

A notification is recorded as accepted and nothing more: it is never counted as a crawl or an index entry. Because Google doesn't use IndexNow, Google is handled separately: the URL Inspection API reports what Google says about each page, and the Chrome extension clicks Request indexing in your own Search Console session, at most 10 pages per property per rolling 24 hours. More on the features page.

Frequently asked questions