OpenAI · Training
What is GPTBot?
OpenAI's crawler for collecting public pages that may be used to train its models. OpenAI says it follows robots.txt and publishes its IP ranges.
- robots.txt token
GPTBot- Operator
- OpenAI
- Purpose
- Training
- Obeys robots.txt
- Yes The operator says it follows robots.txt.
- Documented by operator
- Yes, on the operator's own pages.
- Published IP list
- openai.com/gptbot.json
- Documentation
- platform.openai.com
Blocking GPTBot does not affect ChatGPT search or user-requested fetches, which come from OAI-SearchBot and ChatGPT-User.
Block GPTBot from your whole site
User-agent: GPTBot
Disallow: /
Allow it, but keep private paths out
User-agent: GPTBot
Allow: /
Disallow: /admin/
Other bots from OpenAI
Add it to a robots.txt in the generator →