Bingbot IP Ranges 2026: bingbot.json List and Log Check
Microsoft's bingbot.json lists 29 IPv4 prefixes, no IPv6, created 3 January 2024. Here is how an earlier copy matched 103 Bingbot requests on one site.
- bingbot
- bing
- ip ranges
- bot verification
- reverse dns
- server logs
Microsoft publishes the Bingbot IP ranges at bing.com/toolbox/bingbot.json: 29 IPv4 prefixes, no IPv6, with a creationTime of 3 January 2024. In our server logs from 19 to 31 October 2026, 97 of 103 Bingbot requests came from inside the 28-prefix copy of that file we fetched on 1 November, and the 6 that did not all used one short user agent.
As of 9 November 2026. The log counts below were made with the 28-prefix copy from 1 November.
Picture three lines in your access log. All three say Bingbot, but the user agents differ. These lines are invented for this article and use documentation-range addresses:
203.0.113.10 - - [31/Oct/2026:10:00:01 +0000] "GET /robots.txt HTTP/1.1" 200 120 "-" "Mozilla/5.0 AppleWebKit/537.36 (KHTML, like Gecko; compatible; bingbot/2.0; +http://www.bing.com/bingbot.htm) Chrome/116.0.1938.76 Safari/537.36"
203.0.113.11 - - [31/Oct/2026:10:00:04 +0000] "GET /guides/ HTTP/1.1" 200 5120 "-" "Mozilla/5.0 AppleWebKit/537.36 (KHTML, like Gecko; compatible; bingbot/2.0; +http://www.bing.com/bingbot.htm) Chrome/116.0.1938.76 Safari/537.36 AppleWebKit/537.36 (KHTML, like Gecko; compatible; bingbot/2.0; +http://www.bing.com/bingbot.htm) Chrome/116.0.1938.76 Safari/537.36"
198.51.100.23 - - [31/Oct/2026:10:00:09 +0000] "GET /wp-login.php HTTP/1.1" 404 150 "-" "Mozilla/5.0 (compatible; bingbot/2.0; +http://www.bing.com/bingbot.htm)"
A user agent is text that any client can send. The client address decides, and Microsoft documents two ways to test it. Where does the list live, and what does Microsoft say about it?
Where are the Bingbot IP ranges published?
Microsoft serves the Bingbot IP ranges as a JSON file at https://www.bing.com/toolbox/bingbot.json. On 9 November 2026 it held 29 prefixes under the key ipv4Prefix, no IPv6 entry, and a creationTime of 2024-01-03T10:00:00.121331. Fifteen of the prefixes are /28 blocks, and together all 29 cover 4,752 addresses. On 1 November it held 28 prefixes covering 4,736 addresses.
Microsoft's Bing Webmaster help pages (read 9 November 2026) describe three checks and five crawler names. This table is built from those pages only:
| Question | What Microsoft's page says | Page |
|---|---|---|
| Is there an IP list? | Match the address "to the list of Bingbot IP addresses". "Refresh this list daily as we can change this list any time." | Verifying Bingbot |
| Reverse DNS | The address "resolves to a name that end with search.msn.com", and a forward lookup must confirm "that it resolves back to the same IP address". | Verifying Bingbot |
| Verify Bingbot tool | It "allows you to check if an IP address belongs to Bingbot." It sits in Bing Webmaster Tools under "Tools & Enhancements" and as a "public tool at http://www.bing.com/toolbox/verify-bingbot". | Verify Bingbot |
| Crawler names | Bingbot, AdIdxBot, BingPreview, MicrosoftPreview and BingVideoPreview. User agent strings "are easy to spoof". | Overview of Bing crawlers |
| Which crawlers use the list | Not stated on the three pages. | none |
| IPv6 | Not mentioned on the three pages. The file has no IPv6 prefix. | none |
The "none" rows are a reading of three pages on one day, not a statement by Microsoft. One row matters for your logs: the crawler page lists the same desktop string, bingbot/2.0 with a Chrome version, for both Bingbot and BingPreview. A bingbot/2.0 token alone therefore does not tell you which of the two sent the request.
So does a real site's Bingbot traffic match this file?
Do real Bingbot requests match the list? Our log numbers
On this site, 97 of 103 Bingbot requests between 19 and 31 October 2026 came from addresses inside the 28-prefix copy of bingbot.json. Both common user agent forms matched completely. The 6 requests outside the file shared one address and one short user agent, which is worth checking but not proof of an impostor.
Method. Period: 19 to 31 October 2026, UTC. Sample: 103 requests in this site's raw bot archive whose user agent contains "bingbot" in any letter case, 21 of them fetching robots.txt. Counting: each client address was tested against the 28 prefixes in the file we fetched on 1 November 2026, and requests were grouped by normalized user agent. Limits: one site, 13 days, 103 requests. The archive keeps only requests classified as bots. The list is a snapshot. Our read API's bot counter shows 58 Bingbot requests for the same period, because it counts differently from this substring match, and we use the 103 here. The raw access logs for that window have since been rotated out, so we could not rerun the count against the current 29-prefix file. On the logs still available (2 to 9 November 2026), all 14 requests to this site with a Bingbot user agent came from addresses inside the 29 prefixes.
| User agent form | Requests | Distinct addresses | Address in Microsoft's list |
|---|---|---|---|
| Standard: Chrome version suffix | 57 | 37 | yes, all |
| Doubled: the bingbot part twice in one string | 40 | 16 | yes, all |
Short: Mozilla/5.0 (compatible; bingbot/2.0; +http://www.bing.com/bingbot.htm) |
6 | 1 | no, all outside |
Two findings stand out. First, the doubled string is odd, and Microsoft's Bingbot row does not list it. Yet all 40 requests came from 16 listed addresses. An unusual string from a listed address is Bingbot behaving oddly, and the address, not the string, makes the call.
Second, the short string is not an invention of ours: Microsoft's August 31, 2012 post uses it as its example, and the current crawler page lists it as one of three Bingbot strings. So the format alone proves nothing. What stands out is that the single address behind all 6 short requests is outside the file. A new crawl machine could be missing from a list that Microsoft says changes at any time, and we ran no reverse lookup on that address for this article. Treat it as unverified.
Our fake bot study shows the same direction for 2 to 8 November 2026: of 39 requests claiming to be Bingbot, 36 came from the published ranges and 3 did not (92.3 %), and 94.6 % over the newest three weeks. The Bingbot profile has the robots.txt token and the operator links.
If the file is the test, how much should its 2024 date worry you?
Is a creationTime of 3 January 2024 a sign the list is stale?
Not by itself. The creationTime in bingbot.json was 1,041 days old on 9 November 2026, yet the file's HTTP Last-Modified header was recent and kept moving: it read Fri, 23 Oct 2026 in our first fetch on 1 November, Thu, 29 Oct 2026 in a later one, Sun, 01 Nov 2026 15:16:45 GMT when we fetched again at 20:05 UTC, and Wed, 04 Nov 2026 13:25:42 GMT on 9 November. Through 1 November the creationTime, the prefix count (28) and the address total (4,736) were the same at every reading. On 9 November the file held 29 prefixes and 4,752 addresses, one more /28 block than on 1 November, with the same creationTime. Microsoft's own page tells you to "Refresh this list daily as we can change this list any time."
Read the two dates for what they are. creationTime is a field inside the JSON, and Microsoft does not say what it measures. It did not move when a prefix was added. Last-Modified is an HTTP header. Its first three readings came with the same prefix count and address total, so we cannot tell whether any prefix changed on those dates. The 4 November reading came with one more prefix than we counted on 1 November, but we kept no earlier copy and cannot say which prefix is new. The safe habit is to fetch the file on your own schedule, keep the last copy and compare the set of prefixes, not the date.
Older advice adds noise. The 2012 post says: "Bing does not publish a list of IP addresses or ranges from which we crawl the Internet." That statement is superseded, since the current Verifying Bingbot page points to a list. The same post also describes the reverse and forward DNS test, and the current page still gives that test.
How do you run the list check and the DNS check on your own log?
How do you check an address against the list and with DNS? A tested run
Parse each ipv4Prefix from bingbot.json with Python's ipaddress module and test the client address with in. We ran the script below against Microsoft's live file on 1 November 2026 and updated the outputs on 9 November 2026. It prints INSIDE or OUTSIDE for each Bingbot line in a combined-format log, where the client address is the first field and the user agent the last quoted field.
import ipaddress
import json
import re
import sys
import urllib.request
URL = "https://www.bing.com/toolbox/bingbot.json"
def load():
with urllib.request.urlopen(URL, timeout=20) as r:
data = json.load(r)
modified = r.headers.get("Last-Modified")
nets = [ipaddress.ip_network(p.get("ipv4Prefix") or p.get("ipv6Prefix"))
for p in data["prefixes"]]
return data["creationTime"], modified, nets
def inside(ip, nets):
addr = ipaddress.ip_address(ip)
return any(addr in n for n in nets)
created, modified, nets = load()
print(f"creationTime={created} Last-Modified={modified} prefixes={len(nets)}")
if sys.argv[1] == "selftest":
first = str(nets[0].network_address) # first address of the first listed prefix
for label, ip in (("192.0.2.10 (documentation)", "192.0.2.10"),
("203.0.113.7 (documentation)", "203.0.113.7"),
("first address of prefix #1", first)):
print(f"{label:<28} inside={inside(ip, nets)}")
else:
# combined log format: client IP first, user agent is the last quoted field
line_re = re.compile(r'^(\S+) .*"([^"]*)"$')
for line in open(sys.argv[2]):
m = line_re.match(line.rstrip("\n"))
if m and "bingbot" in m.group(2).lower():
ip, ua = m.groups()
print(f"{ip:<15} {'INSIDE ' if inside(ip, nets) else 'OUTSIDE'} {ua}")
The self-test uses two documentation addresses, which can never be in Microsoft's list, and the first address of the file's first prefix, which must be:
$ python3 -I bingbot_check.py selftest
creationTime=2024-01-03T10:00:00.121331 Last-Modified=Wed, 04 Nov 2026 13:25:42 GMT prefixes=29
192.0.2.10 (documentation) inside=False
203.0.113.7 (documentation) inside=False
first address of prefix #1 inside=True
Then the three-line sample log, with the browser line skipped and both invented Bingbot addresses reported as outside:
$ python3 -I bingbot_check.py scan sample.log
creationTime=2024-01-03T10:00:00.121331 Last-Modified=Wed, 04 Nov 2026 13:25:42 GMT prefixes=29
203.0.113.10 OUTSIDE Mozilla/5.0 AppleWebKit/537.36 (KHTML, like Gecko; compatible; bingbot/2.0; +http://www.bing.com/bingbot.htm) Chrome/116.0.1938.76 Safari/537.36
198.51.100.23 OUTSIDE Mozilla/5.0 (compatible; bingbot/2.0; +http://www.bing.com/bingbot.htm)
The second method needs two lookups, which Microsoft's page describes for Windows and Linux. On Linux:
host <address-from-your-log> # expect a name ending in search.msn.com
host <name-from-the-first-step> # expect the same address back
We tested this on the file itself: for each of the 29 prefixes we looked up the first usable address on 9 November 2026. Reverse names ended in .search.msn.com for 27 prefixes, and the forward lookup returned the same address all 27 times. The other 2 prefixes had no reverse name at the address we tried, so one address per prefix is a spot check, not a proof of every address. A documentation address has no reverse name at all: host answered "not found: 3(NXDOMAIN)".
What should you do when a request fails both tests?
What should you do with a Bingbot request from outside the list?
Treat it as unverified, then run the reverse and forward DNS test or Microsoft's Verify Bingbot tool. A miss on the list alone is no reason to block, because Microsoft says it can change the list at any time. A miss on the list and on DNS is a sound reason to rate-limit or block that address.
- Keep the user agent form, the path and the address of the request.
- Fetch a fresh copy of the file and test again.
- Run the two
hostlookups, or enter the address in the Verify Bingbot tool. - If both fail, restrict that address. If only the list fails, watch it before you act.
Allowlisting by address needs the same care, and the method is the same as for GPTBot IP ranges. In 2012, Microsoft warned that new Bingbot crawl machines with new addresses "may unintentionally be denied access to your pages" when a server allows Bingbot by a fixed list. Refresh the copy daily, as the current page asks.
The takeaway: the address decides, and two independent tests beat one. Next step: paste any address from your log into the Bingbot option of the crawler IP checker, and look up odd user agent strings with the user agent lookup.
Common follow-up questions
How do I verify that a request is really Bingbot?
Use two independent tests. Match the client address against the prefixes in bingbot.json, and run a reverse lookup that must end in search.msn.com with a forward lookup returning the same address. Microsoft also offers a Verify Bingbot tool in Bing Webmaster Tools.
Is there an official Bingbot IP list?
Yes. Microsoft serves one at https://www.bing.com/toolbox/bingbot.json. It held 29 IPv4 prefixes on 9 November 2026. The Verifying Bingbot page points to it and asks you to "Refresh this list daily as we can change this list any time."
Does Bingbot use IPv6 addresses?
The published file has no IPv6 prefix, and the three Bing pages we read do not mention IPv6. That tells you what Microsoft lists, not whether Bingbot never connects over IPv6. If an IPv6 address claims to be Bingbot, run the reverse DNS test.
What is the difference between Bingbot and BingPreview in a log?
BingPreview "generates page snapshots for Bing" and uses the same desktop bingbot/2.0 string as Bingbot in Microsoft's table. AdIdxBot, MicrosoftPreview and BingVideoPreview carry their own names. Microsoft does not say on the pages we read whether they share the Bingbot list.
What should a Bingbot reverse DNS name end with?
Microsoft's Verifying Bingbot page says the reverse lookup should resolve to a name that ends with search.msn.com, and the forward lookup of that name must return the same address. The page spells it "end with". In our spot check, 27 of 29 listed prefixes passed both lookups.
Sources
- Bingbot IP ranges (JSON), https://www.bing.com/toolbox/bingbot.json, Last checked: 9 November 2026
- Verifying Bingbot (Bing Webmaster help), https://www.bing.com/webmasters/help/verifying-that-bingbot-is-bingbot-3905dc26, Last checked: 9 November 2026
- Verify Bingbot (Bing Webmaster help), https://www.bing.com/webmasters/help/verify-bingbot-2195837f, Last checked: 9 November 2026
- Overview of Bing crawlers (Bing Webmaster help), https://www.bing.com/webmasters/help/which-crawlers-does-bing-use-8c184ec0, Last checked: 9 November 2026
- How to Verify that Bingbot is Bingbot (Bing Webmaster Blog, 31 August 2012), https://blogs.bing.com/webmaster/2012/8/How-to-Verify-that-Bingbot-is-Bingbot/, Last checked: 9 November 2026
- Fake bot study (this site), /data/fake-bots, Last checked: 9 November 2026
- Bingbot request counts and list test, 19 to 31 October 2026 (this site's server logs, own data)
Checked 9 November 2026.