The addresses bots use, by operator.
Ranges that search engines', AI companies' and SEO tools' crawlers, uptime monitors, internet scanners and webhook senders send their requests from, each with its operator, the bots it runs and whether the operator vouches for the address.
- BotopenaiHigh20.125.66.80/28Last seen 2026-11-21
- BotappleHigh17.58.101.36Last seen 2026-11-16
- BotanthropicHigh34.152.124.140Last seen 2026-11-21
- BotcensysHigh66.132.153.63 – 66.132.153.65Last seen 2026-11-20
- Entries
- 39,651
- Refresh
- Daily
Schema
See documentation →Published as CSVGZ and MMDB. Every column is described in the docs.
| start_ip | end_ip | provider | agents | categories | identity | verified_by | confidence | first_seen | last_seen |
|---|---|---|---|---|---|---|---|---|---|
| 20.125.66.80 | 20.125.66.95 | openai | GPTBot | ai_training | verified | list | high | 2026-11-21 | 2026-11-21 |
| 17.58.101.36 | 17.58.101.36 | apple | Applebot | search_engine | verified | rdns | high | 2026-11-16 | 2026-11-16 |
| 34.152.124.140 | 34.152.124.140 | anthropic | Claude-SearchBot|Claude-User|ClaudeBot | ai_assistant|ai_search|ai_training | verified | list | high | 2026-11-21 | 2026-11-21 |
| 66.132.153.63 | 66.132.153.65 | censys | CensysInspect | scanner | declared | high | 2026-11-20 | 2026-11-20 | |
| 2001:4860:4801:f::f | 2001:4860:4801:f::f | Google-CloudVertexBot|GoogleOther|Googlebot | ai_search|other|search_engine | verified | list | high | 2026-11-19 | 2026-11-21 |
The operators behind the IDs.
provider is an ID that holds from build to build. Join it on provider.
| id | name | url | active | note |
|---|---|---|---|---|
| censys | Censys | https://censys.com/ | t | Runs CensysInspect. |
| https://developers.google.com/crawling | t | Runs Googlebot, GoogleOther, Google-CloudVertexBot, AdsBot-Google, FeedFetcher-Google, Google-Site-Verification, Google-Read-Aloud, Google-NotebookLM, Google-Agent. | ||
| openai | OpenAI | https://platform.openai.com/docs/bots | t | Runs GPTBot, OAI-SearchBot, ChatGPT-User, OAI-AdsBot. |
Downloading it from code.
Database API reference →One call gets you the current Bot IP build. Every official client wraps it three ways: straight to disk, a time-limited link you hand to your own runner, or bytes in memory. A file written to disk lands only once the whole transfer has arrived, and checksums() returns the published digests to verify it against.
download()downloadUrl()downloadBytes()metadata()checksums()What it's used for
Three decisions Bot IP is bought for, and what each needs from the data rather than from a score.
Let the right crawlers in
Pass a search engine's crawler through your VPN and proxy rules by the operator that vouches for its address, never by its user agent alone.
Catch impostors
A request naming Googlebot from an address Bot IP doesn't give Google is posing as it, to get past rules that let crawlers in.
Keep bots out of your numbers
Drop the views, signups and clicks crawlers, monitors and link checkers make before you report or pay on them.
Official SDKs
All SDKs on GitHub →Twelve official server clients wrap the database endpoints as well as the lookup: list what you're licensed for, poll a build, follow the download redirect. Install commands are in the docs.
Common questions
What does a row tell us?
One range or address a bot uses: its operator's ID in Bot Provider, the bots it serves as their user agents name them, what they do, whether the operator vouches for it and how, a confidence, and the first and last day it was seen.
What does verified mean?
The operator vouches for the address: it publishes it as its bot's, or documents the reverse DNS name its bot's addresses carry. A row that isn't verified names a bot that calls itself one, which is a claim rather than a fact.
Will it flag a crawler as a VPN?
No. A bot's address answers is_bot apart from the VPN and proxy flags, so a rule can let a verified search engine through while it blocks VPNs.
Should we block these addresses?
Not wholesale. A search engine's crawler brings you visitors, where a training crawler takes your content: decide per operator and per category, and treat a request that names a verified bot from an address it doesn't use as an impostor.
How often is it rebuilt?
Daily. The metadata call says when the current build landed and how many rows it holds, so you fetch only a new one.
Which formats does it ship in?
Gzipped CSV and MMDB, both cut from the same build. The MMDB opens in any MaxMind DB reader, so lookups run in memory on your own servers.
How do we get it?
Under a database license, agreed per dataset and bundled with Enterprise plans. Request access names Bot IP to us, and once it's licensed it downloads from the console, the database API or any official client, or each new build is announced to your webhook, or written to your AWS S3 or S3-compatible storage, Google Cloud Storage, Azure Blob Storage, or an SFTP or FTPS server. Download sample gets you rows from the current build first.
Is it the same data the API answers from?
Yes. A lookup's is_bot comes from these builds, so the file and the API agree on an address.
What if a row looks wrong?
Send it to us and we'll tell you what we hold on it and when it was last seen. Corrections land in the next build.
Unlimited lookups, no sign-up.
Paste it into a terminal. Any address answers its VPN, Tor, relay and anycast flags, where it is, which network announces it and its hostname, with no daily or monthly cap.
{"ip": "146.70.116.130","is_bogon": false,"is_vpn": true,"is_relay": false,"is_tor": false,"is_bot": false,"is_anycast": false,"relay": {},"tor": {},"location": {"continent": "EU","continent_name": "Europe","country": "AT","country_name": "Austria","timezone": "Europe/Vienna","currency_code": "EUR","currency_name": "Euro","currency_symbol": "€","flag_emoji": "🇦🇹","flag_unicode": "U+1F1E6 U+1F1F9"},"asn": {"asn": 9009,"name": "M247 Europe SRL","domain": "m247global.com","type": "hosting","prefix": "146.70.116.0/24","country": "RO","rpki_status": "not_found"},"hostname": ""}