Bingbot

Operator

Microsoft

User agent

Mozilla/5.0 (compatible; bingbot/2.0; +http://www.bing.com/bingbot.htm)

Robots token

bingbot

Obeys robots.txt

Yes, per the operator

Verification

Reverse DNS lookup. Reverse DNS must resolve to a hostname ending in search.msn.com, and a forward lookup on that hostname must return the same IP. Microsoft also publishes CIDR ranges at bing.com/toolbox/bingbot.json.

Operator docs

https://www.bing.com/webmasters/help/which-crawlers-does-bing-use-8c184ec0

The documented Bingbot string:

Mozilla/5.0 (compatible; bingbot/2.0; +http://www.bing.com/bingbot.htm)

Bing also crawls from a mobile variant, which wraps a mobile device string around the same bingbot/2.0 token. Because the device half of that string changes without notice, write your rules against the bingbot substring and ignore the rest.

The robots.txt token is bingbot. Token matching is case-insensitive, so Bingbot and bingbot are the same rule.

Who operates it and what it feeds

Microsoft. The Bing index is a bigger dependency than its search market share suggests: Copilot answers read it, and several other search products license it. Blocking Bingbot is therefore a wider decision than blocking one search engine.

Does it honor robots.txt

Yes, and Bing supports two things Google does not. Crawl-delay is read and respected, which gives you a rate limit without writing any code. Bing Webmaster Tools also offers a crawl control schedule with per-hour rates, useful when your origin is fine at 3am and struggling at noon.

As with any crawler, a Disallow rule prevents the fetch, so a noindex tag on a disallowed URL is never read. Pick one mechanism per URL.

How to verify a request

Microsoft's published check is a forward-confirmed reverse lookup:

host 157.55.39.1
# expect a hostname ending in search.msn.com
host msnbot-157-55-39-1.search.msn.com
# expect the original IP back

Bing's own guidance is blunt about why the DNS check comes first: the address ranges change at any time, so a hardcoded IP allowlist goes stale. Microsoft does publish the ranges as JSON at https://www.bing.com/toolbox/bingbot.json for bulk log analysis, and Bing Webmaster Tools includes a Verify Bingbot tool that takes an IP and answers directly. Use the JSON for batch work and DNS for anything live.

Allow it

User-agent: bingbot
Allow: /

Add a rate cap if your origin is thin:

User-agent: bingbot
Crawl-delay: 1
Allow: /

Block it

User-agent: bingbot
Disallow: /

What blocking costs you

Bing organic traffic, plus Copilot citations, plus whatever downstream products license the index. For most sites that lands somewhere between three and ten percent of search traffic, but the distribution is lumpy. Desktop-heavy, enterprise and North American audiences over-index on Bing, and if your traffic skews that way the loss is larger than the market-share number predicts.

The more common Bingbot problem is not blocking, it is volume. Bing crawls some sites harder than Google does, and a thin origin can start returning 503 or 429 under it. Reach for Crawl-delay or the crawl control schedule before you reach for Disallow, and check time to first byte while you are there, since slow responses make any crawler look heavier than it is.

If Bingbot is fetching a lot and indexing little, look at soft 404 responses and duplicate canonical targets first. Bing, like Googlebot, spends its budget on whatever URLs you expose, and a faceted URL space will absorb an arbitrary amount of it.

Find it in your logs

awk '$0 ~ /bingbot/ {print $1}' access.log | sort | uniq -c | sort -rn | head

Run the reverse and forward lookups on the top addresses before you read anything into the counts. Failures are common, and a large block of unverified bingbot traffic from one network is usually a scraper that picked the string because sites allow it.

A caveat specific to Bing

Bing's index has a second consumer that is easy to forget. Copilot answers read it, so a Disallow aimed at keeping your content out of an AI product also removes you from ordinary Bing web results, and there is no token that separates the two. Microsoft offers no equivalent of Google-Extended or Applebot-Extended. If the AI use is your objection, the only available lever takes the search traffic with it.

Bing Webmaster Tools reports crawl errors per URL, which is worth checking before you change any rule. A rise in 403 or 404 responses to bingbot is nearly always a server or CDN change rather than a robots.txt one, and editing the file will not fix it.

Check your site against Bingbot