Crawler IP ranges
Every crawler operator that publishes the addresses its bots use — Google, Microsoft, OpenAI, Perplexity, DuckDuckGo and Ahrefs — and the reverse DNS check for Apple, Yandex and Baidu, which publish none. The ranges come from the operators' own files, synced daily by the same job that feeds the IP lookup, and each page prints the date of its copy.
Every crawler the lookup knows
A User-Agent string is a claim; an address range is proof. The table lists each crawler with what it does, who operates it, how many prefixes its published file holds today and when the copy here was last synced. A crawler whose file is not synced on this server still has a page, marked as such.
| crawler | kind | operator | prefixes | synced |
|---|---|---|---|---|
| Googlebot | search | 170 IPv4 · 147 IPv6 | 2026-09-12 | |
| Google special crawlers | fetcher | 136 IPv4 · 136 IPv6 | 2026-09-12 | |
| Google user-triggered fetchers | fetcher | 529 IPv4 · 529 IPv6 | 2026-09-12 | |
| Google user-triggered fetchers | fetcher | 248 IPv4 · 248 IPv6 | 2026-09-12 | |
| Bingbot | search | Microsoft | 28 IPv4 · 0 IPv6 | 2026-09-12 |
| GPTBot | ai_training | OpenAI | 21 IPv4 · 0 IPv6 | 2026-09-12 |
| ChatGPT-User | ai_assistant | OpenAI | 213 IPv4 · 0 IPv6 | 2026-09-12 |
| OAI-SearchBot | search | OpenAI | 39 IPv4 · 0 IPv6 | 2026-09-12 |
| PerplexityBot | search | Perplexity | 8 IPv4 · 0 IPv6 | 2026-09-12 |
| Perplexity-User | ai_assistant | Perplexity | 4 IPv4 · 0 IPv6 | 2026-09-12 |
| DuckDuckBot | search | DuckDuckGo | 486 IPv4 · 0 IPv6 | 2026-09-12 |
| AhrefsBot | seo | Ahrefs | 81 IPv4 · 0 IPv6 | 2026-09-12 |
| Applebot | search | Apple | reverse DNS | no file |
| YandexBot | search | Yandex | reverse DNS | no file |
| Baiduspider | search | Baidu | reverse DNS | no file |
15 of 15 crawlers answered on this server · newest sync 2026-09-12 · checked 2026-09-13
Why this matters
Sites treat crawlers differently from visitors: search bots get unmetered access and a pass through bot protection, AI crawlers get whatever robots.txt says, and everything else gets the ordinary rate limits. Every one of those decisions rests on knowing who is really asking, and the only evidence a request carries is its source address. The IP lookup answers it for one address at a time; these pages are the ranges themselves, for anyone who would rather check locally.
Common questions about crawler ranges
Why verify a crawler by its IP address at all?
Because the User-Agent header is a string anyone can type. A scraper that calls itself Googlebot gets the allowances sites give Googlebot — crawl budget, unmetered access, a pass through bot protection — unless the site checks where the request actually came from. Each operator publishes the address ranges its crawler uses, or supports forward-confirmed reverse DNS, and those are the only two proofs there are.
How current are the ranges on these pages?
They are the operators’ own published files, synced daily by the same job that feeds the IP lookup, and every page prints the date its file was last synced. An operator adds and retires ranges without notice, so a copy pasted into a firewall a year ago is the thing to distrust, not the file.
What does a match prove?
That the request came from that operator’s infrastructure. It says nothing about what the crawler will do with a page: a search indexer, an AI training crawler and a fetcher acting for a user all share the same proof. The crawler_kind field in the API answer separates them, and robots.txt is where a site says which of them are welcome.
Why do some crawlers have no ranges here?
Applebot, YandexBot and Baiduspider publish no address ranges. Their operators document a reverse DNS check instead: look up the PTR name of the address, confirm it ends in the operator’s domain, then resolve that name forward and confirm it points back at the same address. The API does exactly that when a lookup asks for the hostname.
More free tools
All free toolsNeed this at scale?
The Spaw API answers is_crawler, the crawler's name and its kind for any address, from the same ranges and a forward-confirmed reverse DNS check — 10 free lookups a month, no card required.