AI bots
GPTBot: what it is, its IP ranges, and how to block it
GPTBot is OpenAI's crawler for AI training.
| Vendor | OpenAI |
|---|---|
| What it is for | AI training |
| In its User-Agent | GPTBot |
| Its vendor's address list | https://openai.com/gptbot.json |
| IP ranges | 18 |
It takes pages to train AI models and sends nobody back to your site. Bot Protection blocks it by default; you can let it through.
How to verify it
Any bot can write GPTBot in its User-Agent. A request is GPTBot only when its address is in OpenAI's published list below: check the address, not the name.
How to block it
With robots.txt, you ask it to stay out:
User-agent: GPTBot Disallow: /
robots.txt is a request, not a block, and it cannot tell the real crawler from a bot wearing its name. Bot Protection verifies GPTBot by its address on every request and answers it as you choose on the Set-up page, per kind or for GPTBot alone; a bot that only wears the name is judged like any other bot.
Its IP ranges
From OpenAI's own list, downloaded again every day.
IPv4
132.196.86.0/24 172.182.202.0/25 172.182.207.0/25 20.125.66.80/28 20.171.206.0/24 20.171.207.0/24 4.227.36.0/25 52.230.152.0/24 74.7.175.128/25 74.7.227.0/25 74.7.227.128/25 74.7.228.0/25 74.7.230.0/25 74.7.241.0/25 74.7.241.128/25 74.7.242.0/25 74.7.243.128/25 74.7.244.0/25