---
title: "Bingbot: Microsoft’s search crawler, its user agents and how to verify it"
description: Bingbot crawls the web for Bing, the index Copilot searches. Its user agents, reverse DNS verification, crawl-delay, and how it differs from AdIdxBot.
canonical_url: https://rebilder.com/learn/bingbot
last_updated: "2026-11-15"
---

# [Bingbot: Microsoft’s search crawler, its user agents and how to verify it](https://rebilder.com/learn/bingbot)

> Bingbot crawls the web for Bing, the index Copilot searches. Its user agents, reverse DNS verification, crawl-delay, and how it differs from AdIdxBot.

- **Updated:** 2026-11-15
- **Author:** Rebilder
- **Published:** 2026-11-15

**Bingbot is Microsoft’s standard web crawler: it fetches pages for the Bing search index, and Bing’s own documentation describes it as the crawler that handles most of Bing’s daily crawling.** It identifies itself with the `bingbot` token, reads `robots.txt`, honors `Crawl-delay`, and can be verified with a reverse DNS lookup that resolves to `search.msn.com`.

Bingbot matters beyond Bing’s results page. Microsoft documents that Microsoft Copilot grounds answers by sending search queries to the Bing search service, so the pages Bing has crawled are the pages Copilot’s web search can draw on. This guide covers bingbot’s current user agents, how it differs from AdIdxBot, BingPreview and MicrosoftPreview, how to verify a request, how to slow it down, and the controls Bing offers for AI answers.

## What is bingbot?

Bing runs several crawlers, and its Webmaster Tools help page “Which crawlers does Bing use?” names bingbot as the standard one. It discovers URLs through links, sitemaps and submissions, fetches and renders the pages, and feeds what it finds into the Bing index. If a page is not in that index, Bing cannot rank it, and services built on Bing search cannot return it.

- **It reads `robots.txt`.** A `User-agent: bingbot` group controls it, and Bing provides a robots.txt tester in Webmaster Tools.
- **It supports `Crawl-delay`.** Unlike Googlebot, bingbot reads the directive, and Bing Webmaster Tools adds a crawl control setting on top.
- **It renders pages.** Bing’s user agent tracks the Microsoft Edge version its rendering engine uses, which is why the string carries a Chrome version.
- **It can be verified.** Microsoft documents a reverse and forward DNS check, and offers a Verify Bingbot tool for single addresses.

Microsoft’s own pages, for the current details

```text
Which crawlers does Bing use?   https://www.bing.com/webmasters/help/which-crawlers-does-bing-use-8c184ec0
Verify Bingbot tool:            https://www.bing.com/toolbox/verify-bingbot
User-agent change (Bing blog):  https://blogs.bing.com/webmaster/april-2022/Announcing-user-agent-change-for-Bing-crawler-bingbot
Verify bingbot (Bing blog):     https://blogs.bing.com/webmaster/2012/08/31/how-to-verify-that-bingbot-is-bingbot
Copilot web search:             https://learn.microsoft.com/en-us/copilot/plan/manage-public-web-access
```

## Bingbot user agent strings

Bing moved bingbot to “evergreen” user agents and retired the older short string. The current desktop and mobile formats, as Bing published them, look like this:

Bingbot user agents (Bing Webmaster blog)

```http
# Desktop
User-Agent: Mozilla/5.0 AppleWebKit/537.36 (KHTML, like Gecko; compatible; bingbot/2.0; +http://www.bing.com/bingbot.htm) Chrome/W.X.Y.Z Safari/537.36

# Mobile
User-Agent: Mozilla/5.0 (Linux; Android 6.0.1; Nexus 5X Build/MMB29P) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/W.X.Y.Z Mobile Safari/537.36 (compatible; bingbot/2.0; +http://www.bing.com/bingbot.htm)

# Retired short form, still found in old logs and fixtures
User-Agent: Mozilla/5.0 (compatible; bingbot/2.0; +http://www.bing.com/bingbot.htm)
```

`W.X.Y.Z` is a placeholder. Bing substitutes the Microsoft Edge version its rendering engine currently runs, so the number moves as Edge updates. Bing’s own advice is to stop matching the whole string and match the stable parts instead: the `bingbot` token and the `bingbot.htm` URL.

The mobile string is the one to test against if your site serves different markup by device. Bing crawls with both, and a mobile template that hides prices or specifications behind a tap hides them from the mobile crawler too.

## Bingbot vs AdIdxBot vs BingPreview vs MicrosoftPreview

Microsoft lists more than one crawler, and they do different jobs. Treating them as one bot leads to the wrong rule, such as blocking the ads quality crawler while trying to slow down search crawling:

| Crawler | What Microsoft says it does | User-agent token | How Rebilder’s detector classifies it |
| --- | --- | --- | --- |
| **bingbot** | The standard crawler for Bing search; most of Bing’s daily crawling. | `bingbot` | Crawler, `bingbot` platform. Always served canonical HTML. |
| **AdIdxBot** | Crawls ads and follows them to the advertised sites for quality control. | `adidxbot` | Crawler, attributed to the `bingbot` platform. Always served canonical HTML. |
| **BingPreview** | Generates page snapshots for Bing. | `BingPreview` | Not matched by name; classified from its other headers. |
| **MicrosoftPreview** | Generates page snapshots for Microsoft products. | `MicrosoftPreview` | Crawler, its own `microsoft-preview` platform. Always served canonical HTML. |

The practical upshot: AdIdxBot checks the landing pages your Microsoft Advertising ads point to, so it should see the page a shopper sees; blocking it, or serving it something different, works against that review. MicrosoftPreview generates the snapshots Microsoft products show, such as a link preview, so its traffic tells you about sharing, not about search.

AdIdxBot and MicrosoftPreview, as transcribed in Rebilder’s detector fixtures

```http
User-Agent: Mozilla/5.0 (compatible; adidxbot/2.0; +http://www.bing.com/bingbot.htm)
User-Agent: Mozilla/5.0 (compatible; MicrosoftPreview/2.0; +https://aka.ms/MicrosoftPreview)
```

Both have desktop and mobile variants on Microsoft’s crawler page, and the `aka.ms/MicrosoftPreview` link in the second string redirects to that same page. Match on the tokens, not the full strings.

## How to verify bingbot: reverse DNS and the Verify Bingbot tool

Scrapers borrow the bingbot user agent because many sites wave it through. Microsoft’s documented check is a two-step DNS lookup: resolve the source IP to a hostname, confirm the name ends in `search.msn.com`, then resolve that name forward and confirm it returns the same IP.

Reverse, then forward DNS

```bash
# Step 1: reverse lookup on the IP from your logs
host <ip-from-your-logs>
# expect: ... domain name pointer <name>.search.msn.com.

# Step 2: forward lookup on that name
host <name>.search.msn.com
# expect: <name>.search.msn.com has address <ip-from-your-logs>
```

- **Both steps matter.** Anyone can set a reverse DNS record that claims `search.msn.com` for an address they control. Only the forward lookup, which queries Microsoft’s own DNS, proves the name belongs to Microsoft.
- **For a single address, use the tool.** Bing’s Verify Bingbot page takes an IP and tells you whether it is bingbot. It suits spot checks of a suspicious log line.
- **For a firewall allowlist, use the published ranges.** Microsoft serves a JSON list of bingbot IPv4 prefixes on bing.com. Read its `creationTime` field when you load it, and keep the DNS check as the deciding test for addresses outside the list.
- **Cache the result.** Look an IP up once and remember the answer, rather than running DNS on every request.

Bingbot IP prefixes served by Microsoft

```bash
curl -s https://www.bing.com/toolbox/bingbot.json
# {
#   "creationTime": "...",
#   "prefixes": [ { "ipv4Prefix": "..." }, ... ]
# }
```

## Controlling bingbot in robots.txt

Bingbot follows the most specific group that names it. Give it a `User-agent: bingbot` group and it reads that group alone; leave it out and bingbot follows your `User-agent: *` rules. The exception is `Crawl-delay`, covered in the next section, which bingbot reads from either place.

robots.txt: keep bingbot out of private paths only

```text
User-agent: *
Disallow: /account/
Disallow: /checkout/

User-agent: bingbot
Disallow: /account/
Disallow: /checkout/
Disallow: /search?

Sitemap: https://example.com/sitemap.xml
```

Notice the repetition. Once bingbot has its own group, the `*` rules no longer apply to it, so any path you still want excluded has to appear again. Forgetting this is how a site that only meant to add one bingbot rule ends up opening its account pages to Bing.

- **Block the ads crawler separately, if at all.** AdIdxBot has its own token. Bing’s robots.txt tester lets you switch between Bingbot and AdIdxbot to see how each reads your file.
- **Use page-level tags for index control.** A `Disallow` rule stops bingbot fetching a page. To ask Bing to leave a page out of results, serve `<meta name="robots" content="noindex">` on a page bingbot is allowed to fetch, so it can read the tag.
- **Expect a delay.** Bing has said bingbot keeps directives in memory for a few hours, so a change takes effect after the next fetch of the file.

## Bingbot crawl-delay and Crawl Control

If bingbot is fetching harder than your server likes, you have two levers, and the file wins when they disagree. Bing’s webmaster blog explains that a `Crawl-delay` in `robots.txt` overrides the setting in Bing Webmaster Tools.

robots.txt: slow bingbot down

```text
User-agent: bingbot
Crawl-delay: 5
```

- **The value is a window, not a rate.** Bing describes it as the size of a time window in seconds during which bingbot fetches from your site only once. A larger value means fewer pages per day, which also means slower discovery of new pages.
- **It is read from the bingbot group or the `*` group.** This is the one directive bingbot honors outside the single group it otherwise follows.
- **Each subdomain is separate.** A delay in `shop.example.com/robots.txt` governs that host only.
- **Crawl Control shapes the day.** In Bing Webmaster Tools, Crawl Control lets you set crawl rates by hour, so bingbot works harder overnight and lighter during your busiest trading hours, without a blanket delay.

> **Prefer shaping to throttling** A heavy `Crawl-delay` slows every fetch, including the ones that pick up a price change or a new product. If load is the issue at specific times, hourly Crawl Control keeps bingbot current without hammering you at peak.

## Bingbot, Copilot and AI answers

Microsoft’s documentation for Microsoft Copilot and Microsoft Copilot Chat explains how web answers work: when web search is on, Copilot turns the user’s prompt into a short search query, sends it to the Bing search service, and composes its response from what Bing returns. That page names no separate crawler for this. The content Copilot can ground on is the content Bing search can return, and bingbot is what puts it there.

So the decision about bingbot is also a decision about Copilot’s web answers. Block bingbot in `robots.txt` and Bing stops fetching your pages, which leaves Bing, and Copilot’s web search through it, without current content from your site. If you want to stay in Bing search but limit how your content is used in AI answers, Bing offers page-level controls instead:

| Tag | Use in AI chat answers | Use in training Microsoft’s generative AI foundation models | Bing search results |
| --- | --- | --- | --- |
| `<meta name="robots" content="nocache">` | May be included, showing only the URL, title and snippet. | Only URLs, titles and snippets may be used. | Still appears. |
| `<meta name="robots" content="noarchive">` | Not included and not linked. | Not used. | Still appears. |

Bing announced these controls on its webmaster blog when its chat product was called Bing Chat, and said that a page carrying both tags is treated as `nocache`. You can target Bing alone with `<meta name="bingbot" content="nocache">`. Bing’s current webmaster guidelines carry the up-to-date wording for Copilot.

## Help bingbot find changes faster: sitemaps and IndexNow

Bing’s webmaster blog recommends two channels together: sitemaps for complete coverage and IndexNow for URL-level updates the moment they happen. Bing says it fetches a submitted sitemap right away and then revisits it at least daily, and it wants `<lastmod>` to show when the page content actually changed, not when the sitemap file was generated.

1. Reference your sitemap or sitemap index in `robots.txt` with an absolute URL. The details are in [robots.txt sitemap](/learn/robots-txt-sitemap).
2. Add and verify the site in Bing Webmaster Tools, then submit the sitemap under **Sitemaps** so you get fetch status and error reports.
3. Send changed URLs through IndexNow. Bing participates, and submissions are shared with the other participating engines.
4. Use the URL Inspection live test in Webmaster Tools to see a page as bingbot’s desktop user agent fetches it.

## How Rebilder treats bingbot

Rebilder’s gateway answers AI agents with markdown rendered from the sources you connect, such as your catalog and policy pages. Bingbot is not one of those agents. The open-source detector, `@rebilder/agent-detect`, keeps every Microsoft fetcher it knows in its crawler list, and the gateway serves that list your canonical HTML:

- **`bingbot` and `adidxbot`** classify as crawlers on the `bingbot` platform. AdIdxBot is grouped with bingbot because its user agent points at Bing’s bingbot page.
- **`MicrosoftPreview`** classifies as a crawler on its own `microsoft-preview` platform, so link previews in Microsoft products do not inflate your Bing figures.
- **The guardrail is unconditional.** A request matching a crawler token gets canonical HTML even if it sends `Accept: text/markdown`. Search and ad-quality crawlers see exactly the page a customer sees, which is what both Bing’s index and an ad review need. See [Classification](/docs/classification) for the full order of checks.
- **You can see it in the Console.** In **AI visits**, filter the visitor list to crawlers: bingbot rows read **Crawler** under Kind and **Regular page** under Served.

The detector reads headers and does not check source IPs, so it records a claimed bingbot as bingbot. When it matters who really sent a request, run the DNS check above at your CDN or server.

## What is bingbot?

Bingbot is Microsoft’s standard web crawler for Bing search. It fetches and renders pages for the Bing index, identifies itself with the `bingbot` user-agent token, follows `robots.txt` and supports `Crawl-delay`.

## Is bingbot safe to allow?

Genuine bingbot is Microsoft’s search crawler and follows your `robots.txt`. The risk is impostors using its user agent. Confirm a request with a reverse DNS lookup that ends in `search.msn.com` followed by a forward lookup that returns the same IP, or check the address in Bing’s Verify Bingbot tool.

## How do I block bingbot?

Add `User-agent: bingbot` followed by `Disallow: /` to your `robots.txt`. That stops Bing crawling the site, which removes your current content from what Bing search, and Copilot’s web search through it, can return. To limit AI use while staying in search, use the `noarchive` or `nocache` robots meta tags instead.

## Does bingbot respect crawl-delay?

Yes. Bingbot reads `Crawl-delay` from its own group or the `*` group, treats the value as a time window in which it fetches once, and lets it override the crawl setting in Bing Webmaster Tools. Crawl Control in Webmaster Tools can also shape crawling by hour.

## Does Copilot have its own crawler?

Microsoft’s Copilot web search documentation describes Copilot sending generated queries to the Bing search service and building its answer from the results; it names no separate crawler. Allowing bingbot is how your pages get into the index that search draws on.

## What is the difference between bingbot and AdIdxBot?

Bingbot crawls for Bing search. AdIdxBot crawls ads and follows them to advertisers’ landing pages for quality control. They have separate user-agent tokens and separate `robots.txt` groups, so a rule for one does not apply to the other.

## What is MicrosoftPreview?

A Microsoft crawler that generates page snapshots for Microsoft products, such as the preview shown when a link is shared. It is not bingbot, and blocking it affects previews rather than Bing search.

## Why does the bingbot user agent contain a Chrome version?

Bing’s rendering engine is based on Microsoft Edge, and the evergreen user agent advertises the Edge version in use, written in a `Chrome/` token. The version changes over time, so match the `bingbot` token rather than the full string.

## Related

- [robots.txt sitemap: syntax and submission](https://rebilder.com/learn/robots-txt-sitemap)
- [AI agent user agents: the full reference](https://rebilder.com/learn/ai-agent-user-agents)
- [Applebot and Applebot-Extended](https://rebilder.com/learn/applebot)
- [How agent detection works](https://rebilder.com/docs/classification)
- [Which agents Rebilder detects](https://rebilder.com/help/agent-traffic/which-agents-detected)
- [Free AI robots.txt generator](https://rebilder.com/tools/robots-txt-generator)