LLM SEO, also called LLM optimization or LLMO, is the work of making large language models such as ChatGPT, Claude, Gemini and Perplexity's models mention your brand and cite your pages when people ask about your category.

How do LLMs know about my brand?

Two ways. From training data, the text a model learned from before it was released, and from retrieval, when the assistant searches the web while answering. You can influence retrieval directly and quickly; training is slow and indirect.

Should I block GPTBot and ClaudeBot?

That is a business choice about model training. Blocking them does not block search: OpenAI and Anthropic use separate bots (OAI-SearchBot, Claude-SearchBot) to surface sites in answers. Keep those allowed if you want to be cited.

Does llms.txt help LLM SEO?

It may help agents that read it; llms.txt is a proposal for a file that helps language models use a website. Google Search ignores it. It is cheap to add, but it is not a ranking switch.

What content do LLMs cite?

Pages that answer the question directly with specific, checkable facts. In the GEO research paper, adding citations, quotations and statistics improved visibility in generative engine answers, while keyword stuffing did not.

How do I track LLM SEO?

Ask the same buyer questions on each assistant on a schedule and record whether you are named, your position, rivals named instead and the pages cited. One answer on one day tells you little because answers vary.

LLM SEO: How to Get Mentioned by ChatGPT, Claude and Gemini

What is LLM SEO?

Short answerLLM SEO (LLM optimization, or LLMO) is making large language models mention your brand and cite your pages when people ask them about your category. It targets assistants such as ChatGPT, Claude, Gemini and Perplexity rather than a list of blue links.

It overlaps almost entirely with generative engine optimization and AI SEO. The name puts the emphasis on the model: what it already knows about you, and what it finds when it searches.

Two ways an LLM learns about you

Training data

Text the model learned from before release. Slow to change, indirect, and you cannot see inside it. Your lever is being described accurately and often across the web over time.

Retrieval at answer time

The assistant searches the web, reads a few pages and cites them. Fast to influence: let its search bot in and publish the page the answer needs.

Buying questions ("best X for Y", "X vs Y", "X pricing") need current facts, which is when assistants with web search look things up. That is why retrieval is where LLM SEO pays back first.

A retrieved answer: the brand's own page is source 1. Example data · sample brand.

The bots that decide what LLMs can read

Every major assistant separates its search crawler from its training crawler. That lets you choose separately:

Company	Search / answers	Training
OpenAI	OAI-SearchBot; ChatGPT-User when a user's question opens a page1	GPTBot1
Anthropic	Claude-SearchBot; Claude-User2	ClaudeBot2
Perplexity	PerplexityBot; Perplexity-User. PerplexityBot is not used to crawl for foundation models.3	None listed
Google	Googlebot for Search and its AI features	Google-Extended; allowing it also makes a site eligible for grounding in Gemini Apps4

Google also announced on 30 Oct 2026 that it plans to support a shared AI training control in robots.txt, based on IETF work, aiming for 2027.4 Company-specific controls keep working alongside it.

User-agent: GPTBot
Disallow: /

User-agent: ClaudeBot
Disallow: /

User-agent: Google-Extended
Disallow: /

User-agent: CCBot
Disallow: /

User-agent: *
Allow: /

OAI-SearchBotChatGPT searchAllowed

ChatGPT-UserChatGPT opening a page for a userAllowed

GPTBotOpenAI model trainingBlocked

PerplexityBotPerplexity searchAllowed

Perplexity-UserPerplexity opening a page for a userAllowed

Claude-SearchBotClaude searchAllowed

Claude-UserClaude opening a page for a userAllowed

ClaudeBotAnthropic model trainingBlocked

Google-ExtendedGemini training and groundingBlocked

CCBotCommon Crawl, used to train many modelsBlocked

The same three files, read with AY Rank's robots.txt rules. Search bots stay in when only training is blocked.

See which of the 12 AI bots your site lets in.Check my robots.txt

Content that LLMs cite

Assistants answer questions, so they cite pages that answer questions. The GEO research paper tested this: methods like adding citations, quotations and statistics lifted visibility in generative engine answers by up to 40%, and keyword stuffing did not help.7

Say what you are in one line
Category, who it is for, what it costs. Put it on your home, about and pricing pages, in the same words.
Answer comparison questions yourself
Fair "X vs Y" and "alternatives" pages give the model a source that knows your product.
Be specific and checkable
Numbers, dates, limits, sources. Vague claims are easy to skip.
Show who wrote it and why
Google's helpful-content questions ask who made a page, how and why; trust matters most.8
Be consistent off-site
Review sites, directories and community threads should describe you the same way your site does.

Where llms.txt fits

llms.txt is a proposal for a file, usually at your site root, that gives language models a short summary and links to your key pages. Its only required part is an H1 with the site's name.6

GoogleGoogle Search ignores llms.txt; it neither helps nor harms visibility there.5

AgentsTools and agents that read it can use your summary and links. It takes minutes to write.

Write one with our free llms.txt generator, then spend the real effort on access and content.

Measuring LLM SEO

LLM answers vary from run to run. Measure a fixed set of buyer questions on every assistant, on a schedule, and watch the trend:

Metric	Question it answers
Named rate	How often does the assistant name us?
Position	First, or fourth of four?
Share of voice	Who is named instead of us, and how often?
Cited pages	Which pages did it read, and do they mention us?
Wording	What does it say about us: "best for", "cheaper option", nothing?

AY Rank runs this every day on ChatGPT, Gemini, Perplexity, Claude, Google AI Overviews and AI Mode, and keeps each answer with its sources. See brand mentions and the AI rank tracker.

Sources

Overview of OpenAI crawlers · OpenAIChecked 30 Oct 2026.
Does Anthropic crawl the web, and how can site owners block the crawler? · Claude Help CenterChecked 30 Oct 2026.
Perplexity crawlers · PerplexityChecked 30 Oct 2026.
Supporting a shared AI training control for websites · Google Search Central Blog30 Oct 2026. Checked 30 Oct 2026.
Optimizing your website for generative AI features on Google Search · Google Search CentralLast updated 10 Jul 2026. Checked 30 Oct 2026.
The /llms.txt file · llmstxt.orgA proposal, not a search engine standard. Checked 30 Oct 2026.
GEO: Generative Engine Optimization · Aggarwal et al., arXiv 2311.09735 (KDD 2024)Abstract. Checked 30 Oct 2026.
Creating helpful, reliable, people-first content · Google Search CentralLast updated 5 Oct 2026. Checked 30 Oct 2026.

LLM SEO: get mentioned by language models

What is LLM SEO?

Two ways an LLM learns about you

The bots that decide what LLMs can read

Content that LLMs cite

Say what you are in one line

Answer comparison questions yourself

Be specific and checkable

Show who wrote it and why

Be consistent off-site

Where llms.txt fits

Measuring LLM SEO

Sources

LLM SEO, answered.

Hear what LLMs say about you.