AI crawler
What is GPTBot?
GPTBot collects public web pages that OpenAI may use to train its generative AI models. Blocking GPTBot keeps your pages out of OpenAI's training crawls.
- Operator
- OpenAI
- What it does
- Collects pages for AI training
- robots.txt token
GPTBot- Verified on rankecho.io, 30 days
- 134 requests
What GPTBot did on rankecho.io
From 17 October to 15 November 2026, Cloudflare verified 134 requests from GPTBot, on 25 of the 30 days. It fetched 40 different pages, read robots.txt 0 times, sitemaps 92 times and llms.txt 0 times. 1 of its requests got an error status (4xx or 5xx).
- Verified requests
| Date | Verified requests |
|---|---|
| Oct 17 | 0 |
| Oct 18 | 0 |
| Oct 19 | 0 |
| Oct 20 | 0 |
| Oct 21 | 20 |
| Oct 22 | 0 |
| Oct 23 | 10 |
| Oct 24 | 4 |
| Oct 25 | 2 |
| Oct 26 | 2 |
| Oct 27 | 3 |
| Oct 28 | 38 |
| Oct 29 | 1 |
| Oct 30 | 2 |
| Oct 31 | 3 |
| Nov 1 | 2 |
| Nov 2 | 2 |
| Nov 3 | 11 |
| Nov 4 | 4 |
| Nov 5 | 2 |
| Nov 6 | 2 |
| Nov 7 | 1 |
| Nov 8 | 2 |
| Nov 9 | 2 |
| Nov 10 | 2 |
| Nov 11 | 2 |
| Nov 12 | 2 |
| Nov 13 | 2 |
| Nov 14 | 2 |
| Nov 15 | 11 |
| Path | Requests |
|---|---|
/ | 2 |
/learn/best-geo-tools | 1 |
/compare/ai-citation-tracker | 1 |
/learn/whats-in-a-geo-fix | 1 |
/learn/competitor-citation-comparison | 1 |
Also in those days, 124 requests used GPTBot's user agent without Cloudflare verifying it, and 87 of those asked for secret files such as /.env or compose.yml, which no real crawler needs.
Source: rankecho.io's Cloudflare logs, visitor requests only, queried 2026-11-16. Cloudflare sampled 50 of the verified requests, so those counts are estimates.
How many sites block GPTBot
3 of 13 sites whose robots.txt RankEcho read in audits from 3 November 2026 to 16 November 2026 block GPTBot from the whole site. The sample is small and grows with every audit, so read it as a first look rather than a rate for the web.
Blocking GPTBot keeps your pages out of OpenAI's training crawls. It does not remove you from ChatGPT search answers: those use OAI-SearchBot and ChatGPT-User, which have their own tokens.
Allow GPTBot
User-agent: GPTBot Allow: /
Block GPTBot
User-agent: GPTBot Disallow: /
Put the lines in the robots.txt file at the root of your domain. A more specific group for GPTBot replaces the rules under User-agent: * for that crawler.
How to check a visit is really GPTBot
OpenAI publishes the IP ranges for each of its crawlers as JSON files linked from its crawler page. A request with the user agent but an IP outside those ranges is not OpenAI. Cloudflare checks this for you and marks real requests as verified bots.
Source: OpenAI: Overview of OpenAI crawlers.
Questions about GPTBot
What is GPTBot?
GPTBot collects public web pages that OpenAI may use to train its generative AI models.
Should I block GPTBot?
Blocking GPTBot keeps your pages out of OpenAI's training crawls. It does not remove you from ChatGPT search answers: those use OAI-SearchBot and ChatGPT-User, which have their own tokens.
How do I know a visit is really from GPTBot?
OpenAI publishes the IP ranges for each of its crawlers as JSON files linked from its crawler page. A request with the user agent but an IP outside those ranges is not OpenAI. Cloudflare checks this for you and marks real requests as verified bots.
How often does GPTBot visit a site?
On rankecho.io, GPTBot made 134 Cloudflare-verified requests from 17 October to 15 November 2026, on 25 of the 30 days. Your site will differ with its size and how often it changes.
How many sites block GPTBot?
3 of the 13 sites whose robots.txt RankEcho read in audits from 3 November 2026 to 16 November 2026 block GPTBot from the whole site. It is a small sample, so read it as a first look, not a web-wide rate.
Does blocking GPTBot stop ChatGPT from citing my site?
No. ChatGPT search finds pages through OAI-SearchBot and fetches them for users through ChatGPT-User. Block GPTBot to opt out of training while staying citable.
Related crawlers
Check what your robots.txt allows
The free crawler policy check reads your robots.txt and says which AI crawlers it lets in. A free audit also asks ChatGPT, Perplexity and Gemini whether they cite you, and marks a missing answer as crawler access when a crawler is blocked.
Facts checked 2026-11-16 against OpenAI's documentation. Log numbers from 17 October to 15 November 2026. Read our crawler log study for the method.
