GPTBot vs Google-Extended: Which AI Crawler Should You Allow or Block?
Complete side-by-side comparison of GPTBot (OpenAI) and Google-Extended (Google): user-agent strings, safety ratings, SEO impact, and copy-paste blocking rules. Updated August 2026.
Quick Verdict
Both GPTBot and Google-Extended carry the same safety rating (Safe). Your decision should come down to whether you want your content used by OpenAI versus Google, and whether each bot's purpose (training, search, or user requests) aligns with your content strategy.
| Attribute | GPTBot | Google-Extended |
|---|---|---|
| Operator | OpenAI | |
| User-agent string | GPTBot |
Google-Extended |
| Category | AI & LLM Bots | AI & LLM Bots |
| Safety rating | Safe | Safe |
| SEO impact if blocked | Low - No SEO ranking impact | Low - No SEO ranking impact |
| Recommendation | Generally safe to allow; provides legitimate crawling value. | Generally safe to allow; provides legitimate crawling value. |
| Official docs | GPTBot docs | Google-Extended docs |
What is GPTBot?
GPTBot is OpenAI's training crawler. The full profile covers the rest.
What is Google-Extended?
Google-Extended is not a crawler. The full profile covers the rest.
GPTBot vs Google-Extended by the numbers
GPTBot is a training crawler and Google-Extended is an opt-out token rather than a crawler. They do different jobs, so a sensible policy often treats them differently.
| Measure | GPTBot | Google-Extended |
|---|---|---|
| robots.txt token | GPTBot | Google-Extended |
| What it does | collects pages to train AI models | collects pages to train AI models |
| Crawler type | a training crawler | an opt-out token rather than a crawler |
| Priority tier | Tier 1 | Tier 1 |
| Follows robots.txt | Yes, per operator | Not applicable (token) |
| Top-10k robots.txt files | 1,008 named (18.1%), 859 blocked (15.4%) | 801 named (14.3%), 723 blocked (13%) |
| Blocked by top 1k / middle / tail | 20.4% / 15.5% / 14.5% | 18.6% / 13.1% / 11.9% |
In our robots.txt study of the Tranco top 10,000 (5,583 valid files, fetched October 20, 2026), GPTBot is blocked more often than Google-Extended: 859 files against 723. That gap shows many site owners already treat the two differently.
Writing rules for both
Because each crawler reads only the group that names it, you can give GPTBot and Google-Extended different answers in the same file. A common pattern, refusing training while allowing search and assistant fetches, looks like this for these two:
User-agent: GPTBot
Disallow: /
User-agent: Google-Extended
Allow: /- One group each. Do not rely on
User-agent: *to cover either crawler if you have named the other; name both. - Copy shared rules. Any
Disallowlines from the*group that should still apply must be repeated in each named group. - Mind the edge. A bot manager can refuse one and admit the other regardless of robots.txt.
- Verify both. The website crawlability test shows how each is treated on your live site.
Test yourself: if you write User-agent: GPTBot and User-agent: Google-Extended on consecutive lines above one Disallow: /, are both blocked?
Yes. Several User-agent lines in a row form one group, and the rules below apply to every name in it.
Test yourself: does blocking GPTBot also block Google-Extended?
No. Each token is matched separately.
Which to allow, by goal
| Your goal | GPTBot | Google-Extended |
|---|---|---|
| You want AI answers to cite you | Optional | Allow |
| You want to stop AI training | Block | No effect |
| Your server is under load | Rate-limit | Rate-limit |
| You need search traffic | No effect | No effect |
Related: how AI engines choose their sources
Block GPTBot if you do not want your content used for training. Set Google-Extended to control how content is used, not whether it is fetched.
The rest of each operator's family
OpenAI also runs ChatGPT Agent, ChatGPT Operator, ChatGPT-User, ChatGPT-User/2.0, OAI-AdsBot. Rules for GPTBot do not cover them. Google also runs AdsBot-Google, AdsBot-Google-Mobile, APIs-Google, Chrome-Lighthouse, FeedFetcher-Google. Rules for Google-Extended do not cover them.
Read the full profiles for both crawlers, linked in the table above, for documentation links, user-agent strings and server-side rules.
Mistakes when writing rules for GPTBot and Google-Extended together
GPTBot is named in 18.1% of the top-site robots.txt files in our study and blocked in 15.4%; Google-Extended is named in 14.3% of the top-site robots.txt files in our study and blocked in 13%. Large sites block GPTBot more often, which usually reflects a decision about training rather than about Google-Extended.
- They are different kinds of agent (a training crawler versus an opt-out token rather than a crawler), so the usual error is giving them the same rule out of habit. Blocking both to stop training only needs the GPTBot line; the other one changes something else entirely.
- They come from different operators (OpenAI and Google), with different documentation and different update cycles. Do not assume a change one of them announces applies to the other.
- Putting
User-agent: GPTBotandUser-agent: Google-Extendedin separate groups and then editing only one. Two groups drift apart over time; keep a comment above each that says why it differs from the other.
Each crawler's own profile lists the mistakes that apply to it alone; this list is only about combining the two.
Why Google-Extended never shows up next to GPTBot in your logs
Google-Extended is a control token from Google, not a crawler: nothing with that name ever requests a page. GPTBot is a a training crawler from OpenAI that does send requests, so it is the only one of the two you can count, verify or rate-limit.
- A firewall rule can act on GPTBot but not on Google-Extended. The token only works as a robots.txt line.
- Blocking Google-Extended does not slow Google's crawling. It changes what Google may do with pages its other crawlers already fetch.
- In our study of top sites, 13% block Google-Extended and 15.4% block GPTBot. Blocking the crawler is the more common choice here.
How GPTBot and Google-Extended moved in our latest scans
Between our scans of October 17, 2026 and October 20, 2026, files naming GPTBot went from 941 to 1,008 (+67) and files blocking it from 789 to 859 (+70), so more sites wrote a rule for GPTBot and more of those rules block it. For comparison, 2026 and October 20, 2026, files naming Google-Extended went from 743 to 801 (+58) and files blocking it from 666 to 723 (+57), so more sites wrote a rule for Google-Extended and more of those rules block it.
Over the 3 days between those scans, blocking of GPTBot moved +70 files and blocking of Google-Extended moved +57. Owners are acting faster on GPTBot, which is the one to recheck first if you have not touched your file this month.
We first recorded GPTBot on March 17, 2026. OpenAI documents it on developers.openai.com, and we recheck that page before each release. We first recorded Google-Extended on March 17, 2026. Google documents it on developers.google.com, and we recheck that page before each release.
Related: writing content AI engines can cite
Among the 104 AI and LLM crawlers in our database, GPTBot sits ninth by priority (Tier 1), level with ChatGPT Agent, ChatGPT Operator, ChatGPT-User, ChatGPT-User/2.0. 81 of the 104 share its safe rating. On the Google-Extended side: Among the 104 AI and LLM crawlers in our database, Google-Extended sits eighth by priority (Tier 1), level with ChatGPT Agent, ChatGPT Operator, ChatGPT-User, ChatGPT-User/2.0. 81 of the 104 share its safe rating.
How top sites treat GPTBot and Google-Extended
From our robots.txt study of the Tranco top 10,000 (5,583 valid files, fetched October 20, 2026):
GPTBot, the most blocked of the GPTBot and Google-Extended tokens: blocked in 15.4% of files (663 with its own group, 196 throughUser-agent: *), more often by the top 1,000 (20.4%) than the long tail (14.5%).Google-Extended, second of the GPTBot and Google-Extended tokens: blocked in 13% of files (519 with its own group, 204 throughUser-agent: *), more often by the top 1,000 (18.6%) than the long tail (11.9%).
For scale, 20.1% of all files block at least one AI training crawler. GPTBot is one of the tokens driving that figure.
More questions about GPTBot and Google-Extended
- Can I allow one and block the other?
- Yes. Give each its own
User-agentgroup with its own rules. Each crawler reads only the group that names it. - Which of the two visits more often?
- It depends on your site, your audience and how often your content changes. Your access logs are the reliable answer; the commands above count both.
- How long until a change applies?
- Crawlers commonly cache robots.txt for up to a day, so allow about 24 hours and confirm in your logs.
- Do I need to update my rules when either operator launches a new crawler?
- Often, yes. A new token is not covered by rules written for the old ones. Check this comparison and both profiles periodically, or rerun the AI crawler tester after any announcement.
Frequently asked questions
What is the difference between GPTBot and Google-Extended?
GPTBot is operated by OpenAI (GPTBot is OpenAI's training crawler. OpenAI says it crawls "content that may be used in training our generative AI foundation models".) while Google-Extended is operated by Google (Google-Extended is not a crawler. It is a robots.txt token: disallowing it tells Google not to use content crawled by Googlebot for Gemini training and grounding.). They use different user-agent strings (GPTBot vs Google-Extended) and must be blocked separately in robots.txt.
Should I block GPTBot or Google-Extended?
Both GPTBot and Google-Extended carry the same safety rating (Safe). Your decision should come down to whether you want your content used by OpenAI versus Google, and whether each bot's purpose (training, search, or user requests) aligns with your content strategy.
Does blocking GPTBot or Google-Extended hurt SEO?
GPTBot: Low - No SEO ranking impact. Google-Extended: Low - No SEO ranking impact. Neither is a traditional search-ranking crawler, so blocking them does not affect Google or Bing rankings, but it may reduce your visibility in AI-generated answers.
How do I block both GPTBot and Google-Extended?
Add both user-agent groups to your robots.txt: "User-agent: GPTBot" with "Disallow: /" and "User-agent: Google-Extended" with "Disallow: /". Remember robots.txt is voluntary; combine with server-level rules for enforcement.
How to block each bot
Block GPTBot in robots.txt
User-agent: GPTBot Disallow: /
Block Google-Extended in robots.txt
User-agent: Google-Extended Disallow: /
Setting GPTBot, Google-Extended and the rest in one go? Tick both in the robots.txt generator (265+ presets), then confirm the result with the AI crawler checker.
Are GPTBot and Google-Extended allowed on YOUR site?
Run a free check to see exactly which AI crawlers can access your website, plus your AI Visibility Score.
Free AI Crawler Access Check