OpenAI · User-triggered fetch
ChatGPT-User: OpenAI's live page fetcher, and how to control it
ChatGPT-User is the user agent OpenAI uses when a person's action in ChatGPT or a Custom GPT needs a web page, such as answering a question with a live visit or calling a GPT Action. It is not an automatic crawler and is not used to decide search inclusion. OpenAI says robots.txt rules may not apply to it, because a user started the request.
- robots.txt token
ChatGPT-User- User agent
Mozilla/5.0 AppleWebKit/537.36 (KHTML, like Gecko); compatible; ChatGPT-User/1.0; +https://openai.com/bot- Operator
- OpenAI
- Type
- User-triggered fetch
- What it feeds
- Live page visits for user-initiated actions in ChatGPT and Custom GPTs, including GPT Actions; it does not crawl the web automatically.
- Follows robots.txt
- Partly. OpenAI says that because these actions are initiated by a user, robots.txt rules may not apply. It points site owners to OAI-SearchBot for search opt-outs.
- How to verify it
- OpenAI publishes the IP ranges ChatGPT-User uses (list)
Should you block ChatGPT-User?
Blocking ChatGPT-User can stop ChatGPT opening your pages when someone asks about them or pastes a link, which usually means a worse answer about you. OpenAI says robots.txt may not apply, so a firewall rule on its published IPs is the dependable control. It does not change ChatGPT search inclusion.
User-agent: ChatGPT-User
Disallow: /
User-agent: ChatGPT-User
Allow: /
Add these lines to /robots.txt at your site root. A group naming ChatGPT-User replaces your User-agent: * rules for it, so repeat any paths you also want it to skip.
Live check
Is ChatGPT-User allowed on your site?
ChatGPT-User and OpenAI's other crawlers
OpenAI moved its crawler docs to developers.openai.com/api/docs/bots, and the old platform address redirects. ChatGPT-User is the one bot on that page that acts for a specific person rather than on a schedule:
| Token | What it does | Honors robots.txt |
|---|---|---|
| ChatGPT-User | Visits a page when a ChatGPT or Custom GPT user's action needs it | "May not apply" |
| OAI-SearchBot | Surfaces and links sites in ChatGPT search | Yes |
| GPTBot | Crawls content that may train foundation models | Yes |
There is also OAI-AdsBot, which only visits pages submitted as ChatGPT ads. For search inclusion and automatic crawling, OpenAI tells site owners to use OAI-SearchBot, not ChatGPT-User.
What blocking ChatGPT-User does, and what it doesn't
When a person asks ChatGPT about a page, pastes a link, or uses a Custom GPT that calls an external application through GPT Actions, the request arrives as ChatGPT-User. Blocking it means ChatGPT may not be able to read your page at that moment, so it answers from whatever else it has. For a business, that is usually the opposite of what you want: the person was already asking about you.
Blocking it does not remove you from ChatGPT search, and it does not affect training. Those are OAI-SearchBot and GPTBot.
How to block or allow ChatGPT-User
You can write a robots.txt group for it:
User-agent: ChatGPT-User
Disallow: /private/
But OpenAI says robots.txt rules may not apply to user-initiated actions, so treat this as a request, not a guarantee. If you need a hard block, deny the addresses in chatgpt-user.json at your firewall or CDN. To make sure it gets through, check that no bot-protection rule challenges those ranges.
Checking that a visit is really ChatGPT-User
OpenAI gives the full string as Mozilla/5.0 AppleWebKit/537.36 (KHTML, like Gecko); compatible; ChatGPT-User/1.0; +https://openai.com/bot. Unlike the GPTBot and OAI-SearchBot strings, which OpenAI labels examples whose version may change, this one is presented as the full user agent; still, match on the token rather than the exact text. The user agent is easy to fake, so compare the source IP with the published list.
The live check above sends a request with the ChatGPT-User token and compares it with a browser request, which shows whether a WAF or CDN is turning it away. It can't use OpenAI's IPs, so IP-based rules may behave differently. The robots.txt checker shows what your file says to each OpenAI token.
Questions
- Does ChatGPT-User obey robots.txt?
- Not reliably. OpenAI's docs say that because these actions are initiated by a user, robots.txt rules may not apply. If you must refuse it, block its published IP ranges at your firewall or CDN.
- Will blocking ChatGPT-User remove me from ChatGPT search?
- No. OpenAI says ChatGPT-User is not used to determine whether content may appear in search; OAI-SearchBot is the token for search opt-outs.
- Is ChatGPT-User crawling my whole site?
- It shouldn't be. OpenAI says it is not used for crawling the web in an automatic fashion; it visits pages when a ChatGPT or Custom GPT user's request calls for it.
- Do GPT Actions show up as ChatGPT-User?
- OpenAI's docs list GPT Actions, where users interact with external applications from ChatGPT, under the ChatGPT-User section. If you host an API that a Custom GPT calls, expect this user agent.
- What should I allowlist so ChatGPT can read my pages?
- Allow the ChatGPT-User token and the IP ranges in chatgpt-user.json. OpenAI calls its ChatGPT-User string the full user agent: ChatGPT-User/1.0 with a link to openai.com/bot.
Related crawlers
Sources
Checked against these pages on 2026-10-18. Operators change their crawlers; check the source before relying on a detail.