Home / AI bot list / OAI-SearchBot
BOT PAGE · OPENAI · REVIEWED 2026-09-13
OAI-SearchBot: what it is, and how to allow or block it
OAI-SearchBot is the crawler OpenAI uses to build the index behind ChatGPT search. It is not GPTBot, and blocking one is not blocking the other. This page gives you the exact token, what OpenAI does and does not say about it, the robots.txt lines, and how to tell whether it has been to your site.
Where this comes from. Every statement about what OAI-SearchBot does is quoted from, or read directly off, OpenAI's own crawler page — developers.openai.com/api/docs/bots, read 2026-09-13. Where that page does not settle something, this page says so rather than filling the gap with something plausible. The token is spelled exactly as it appears on the AI bot list and in the AI Crawler Checker.
THE FACT SHEET
What OAI-SearchBot is, and who runs it
OAI-SearchBot is an OpenAI crawler whose job is search: it visits pages so that ChatGPT's search features can surface and link to them. That makes it an answer route, not a training crawler.
OAI-SearchBot
Run by OpenAI
- PurposeChatGPT search index. OpenAI: “OAI-SearchBot is for search.”
- KindAnswer route
- robots.txtRead and acted on — OpenAI publishes it as a robots.txt token and says a change can take about 24 hours to take effect.
- TrainingNot the training crawler. OpenAI routes the training opt-out through GPTBot. See what it does not say.
- IP rangesopenai.com/searchbot.json
- Sourcedevelopers.openai.com/api/docs/bots
OpenAI documents three visitors that can arrive at a site under its name, and they do three different jobs. GPTBot collects material that may be used to train models. OAI-SearchBot builds the index that ChatGPT search draws on. ChatGPT-User opens a page because a person in ChatGPT asked it to. They look alike in an access log and are opposite decisions in a robots.txt.
OpenAI's own words for this one: “OAI-SearchBot is used to surface websites in search results in ChatGPT's search features.” And for what a block does: “Sites that are opted out of OAI-SearchBot will not be shown in ChatGPT search answers, though can still appear as navigational links.”
Source: developers.openai.com/api/docs/bots, OAI-SearchBot section, read 2026-09-13.The reason a page exists for this one token is the accident it keeps causing. Plenty of sites blocked GPTBot on purpose and OAI-SearchBot by mistake, with one catch-all rule written before the second token existed. The two are independent, and OpenAI says so: “a webmaster can allow OAI-SearchBot in order to appear in search results while disallowing GPTBot.”
SPELLED EXACTLY
The exact user-agent token, and the full string it arrives with
The token you write in robots.txt is OAI-SearchBot: capital O, A, I, a hyphen, capital S, capital B. A user-agent string recalled slightly wrong is worse than no rule at all, because the rule then matches nothing and the file reads as a decision that was never enforced.
OAI-SearchBot
This is the product token a robots.txt group names. Matching is case-insensitive, so oai-searchbot also matches — but copy it as published.
Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/131.0.0.0 Safari/537.36; compatible; OAI-SearchBot/1.4; +https://openai.com/searchbot
OpenAI labels this an example and says the version number may change. The token is what to look for in a log, not the whole string.
Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/131.0.0.0 Safari/537.36; compatible; OAI-SearchBot/1.4; robots.txt; +https://openai.com/searchbot
OpenAI says it may add a robots.txt marker when fetching the robots file, so a site owner can tell those requests apart from page requests even in a log that records no paths.
Also on OpenAI's page, not covered here
OpenAI's crawler page also lists OAI-AdsBot, which it says only visits pages submitted as ads on ChatGPT. It is not one of the twelve tokens the AI Crawler Checker evaluates and this page makes no claim about it beyond that one sentence.
ONLY WHAT THE OPERATOR DOCUMENTS
What OAI-SearchBot does, and what OpenAI does not say
OAI-SearchBot is documented as a search crawler and nothing else. The training question is answered on OpenAI's page through a different token, and the gap that leaves is stated below rather than papered over.
What OpenAI states
It is for search. “OAI-SearchBot is used to surface websites in search results in ChatGPT's search features.”
Blocking it removes you from ChatGPT search answers. “Sites that are opted out of OAI-SearchBot will not be shown in ChatGPT search answers, though can still appear as navigational links.”
It reads robots.txt, with a delay. “For search results, please note it can take ~24 hours from a site's robots.txt update for our systems to adjust.”
It is independent of GPTBot. “Each setting is independent of the others.”
What OpenAI does not say
It does not say, in so many words, that OAI-SearchBot's fetches are never used for training. The sentence about training is attached to GPTBot: “Disallowing GPTBot indicates a site's content should not be used in training generative AI foundation models.” The training opt-out runs through that token.
It does say the two crawls can be merged. “If your site has allowed both bots, we may use the results from just one crawl for both use cases to avoid duplicative crawling.” So allowing both means one visit may serve both purposes; disallowing GPTBot is the documented way to say no to training while staying in search.
It does not say how often it crawls, or what it does with pages it cannot fetch. Nothing on this page should be read as a crawl rate.
GPTBot
Training input“Used to crawl content that may be used in training our generative AI foundation models.” Blocking it is the documented training opt-out.
OAI-SearchBot
Answer route“For search.” Blocking it takes your pages out of ChatGPT search answers. This page.
ChatGPT-User
Fetch on requestOpens a page when a person in ChatGPT asks it to. OpenAI: “Because these actions are initiated by a user, robots.txt rules may not apply.” And: “Please use OAI-SearchBot in robots.txt for managing Search opt outs and automatic crawl.”
The sentence this site will stand behind is narrow. Blocking OAI-SearchBot removes your pages from the index ChatGPT search draws on, and OpenAI says what that costs. Whether an assistant will then name you anyway, from press coverage, directories or a competitor's comparison page, is not something a robots.txt can control — and it is not something this page measures.
THE FILE
How to allow or block OAI-SearchBot in robots.txt
A robots.txt group for OAI-SearchBot is a User-agent: OAI-SearchBot line followed by its own complete set of rules. Two files below cover the two decisions people actually make; change nothing else until you have read the warning under them.
# Findable in ChatGPT search; content not for training. # OpenAI: "a webmaster can allow OAI-SearchBot in order to # appear in search results while disallowing GPTBot". User-agent: OAI-SearchBot Allow: / User-agent: GPTBot Disallow: /
The position most sites mean when they say no training, still findable. Both groups are complete on their own, which is the point of the warning below.
# Out of the ChatGPT search index. OpenAI says this can take # ~24 hours to take effect after the file changes. User-agent: OAI-SearchBot Disallow: /
This does not block GPTBot or ChatGPT-User: each token is its own decision. Add their groups if you mean them too.
Why every group is written in full
A crawler obeys exactly one group. If any group names OAI-SearchBot, the User-agent: * group is ignored entirely for it — every rule you meant it to inherit is gone. Where several groups match, the longest matching user-agent value wins.
So naming this bot to tighten one thing is how sites accidentally loosen everything else. A named group has to repeat every rule you still want.
Do it with the tools instead of by hand
The robots.txt generator writes a complete group per token for exactly this reason. The AI Crawler Checker then reads your real, live file and reports which group OAI-SearchBot actually lands in, with the line that decided it.
After a change, remember OpenAI's own figure: about 24 hours before its systems adjust for search.
REQUEST, NOT LOCK
How to see whether OAI-SearchBot actually visited
robots.txt is a request, not a lock. It tells a well-behaved crawler what you would like; it stops nothing by itself. The only place a visit is recorded is your own server's access log, and that is where to look.
-
Step 1 · your file
Read what your robots.txt actually says
Not what you meant it to say. The AI Crawler Checker fetches your live robots.txt and returns one verdict for OAI-SearchBot — allowed, blocked, or not measured — with the exact rule line and group behind it.
A CDN can serve a stale copy of the file; this site's own robots.txt did that for a day, and every cache-busted check said it was fine.
-
Step 2 · your log
Look for the token in your access log
Paste a day of your server log into the AI Crawler Log Analyzer. It reads in your browser tab and reports which of twelve AI tokens appeared, when, what each asked for and what your server returned. Nothing is uploaded.
A line carrying OAI-SearchBot and a request for /robots.txt is the crawler reading your rules. Lines for pages are the crawl itself.
-
Step 3 · the claim
Check the address, because the string is a claim
A user-agent string is what the visitor says about itself, and any software can send any string. OpenAI publishes the addresses its crawler uses at openai.com/searchbot.json; a request claiming to be OAI-SearchBot from outside those ranges is not OpenAI.
Everything on this page is about the well-behaved half. A visitor that ignores robots.txt shows up in the log regardless, which is why the log is the check and the file is not.
One more line worth drawing: a visit in your log is not a citation in an answer. OAI-SearchBot fetching a page means the page could be in the index; whether ChatGPT then names you when a buyer asks is a separate measurement, and the AI Visibility Checker is the tool that asks.
QUESTIONS
Before you change a live robots.txt
Is OAI-SearchBot the same as GPTBot?
No. GPTBot is the crawler OpenAI documents as collecting content that may be used in training its models. OAI-SearchBot is the crawler OpenAI documents as building the index behind ChatGPT search. They carry different tokens, they are governed by separate robots.txt groups, and OpenAI states that a site can allow one while disallowing the other.
Does blocking OAI-SearchBot remove me from ChatGPT?
OpenAI says sites opted out of OAI-SearchBot will not be shown in ChatGPT search answers, though they can still appear as navigational links. It does not stop ChatGPT naming you from other material it already holds or from other sites that describe you. Blocking removes your site as a source; it does not remove you from the conversation.
Does OAI-SearchBot use my pages for training?
OpenAI's crawler page attaches the training statement to GPTBot, not to OAI-SearchBot, and it does not say in so many words that OAI-SearchBot fetches are never used for training. It does say that if a site allows both bots, one crawl may be used for both purposes. The documented way to decline training is to disallow GPTBot, and that works whether or not OAI-SearchBot is allowed.
How long does a robots.txt change take to apply?
OpenAI says it can take about 24 hours from a robots.txt update for its systems to adjust for search results. A check run five minutes after an edit is checking your file, not their behaviour.
Does blocking OAI-SearchBot affect Google?
OAI-SearchBot is an OpenAI token and has no part in Google Search. Blocking it changes what ChatGPT search can show, and OpenAI states what that is. Googlebot is a different visitor with a different token, and it is deliberately not discussed on this page.
How do I know it obeyed my file?
Not from the file. A robots.txt is a request, and the only record of a visit is your own access log. Look for the token there, and check the requesting address against the ranges OpenAI publishes, because a user-agent string is a claim any software can make.
Every answer above is either a quotation from OpenAI's crawler page, read 2026-09-13, or a statement about how robots.txt and access logs work. None of it is a measurement of your site.
ONE TOKEN, ONE VERDICT
Read your real robots.txt
against OAI-SearchBot.
The AI Crawler Checker fetches your live file and reports what it says to this crawler and eleven others, with the exact rule line and the group that produced it. A row it cannot read comes back as not measured, never as allowed.