September 5, 2026

llms.txt in 2026: What It Is, Whether You Need It, and What It Actually Does

llms.txt has become one of the most discussed files in the new wave of AI SEO. It is often described as “robots.txt for AI”, a new AI-search standard or even a requirement for getting a website into ChatGPT. In 2026, the reality is much more nuanced. Google Search explicitly says it does not use llms.txt for traditional Search, AI Overviews or AI Mode. At the same time, Chrome Lighthouse now checks for the file within its experimental Agentic Browsing category. OpenAI does not list llms.txt as a requirement for ChatGPT Search either. For search discoverability, allowing OAI-SearchBot to access public content is much more directly relevant. llms.txt is therefore not a magic SEO file or a guarantee of AI visibility. It is an optional machine-readable layer that may be useful to AI systems and agents that choose to consume it.

What is llms.txt?

llms.txt is a proposed Markdown format designed to give language models and AI agents a concise overview of a website and links to important resources. A file may be placed at: https://example.com/llms.txt Its purpose is not necessarily to list every URL on the site. Instead, it can provide a curated map of important information. # Example Company > The company develops software for small businesses. ## Key pages - [About](https://example.com/about): Team and company background - [Services](https://example.com/services): Main services - [Pricing](https://example.com/pricing): Current pricing - [Documentation](https://example.com/docs): Technical documentation

Is llms.txt an official web standard?

It is more accurate to describe it as an emerging convention and open proposal. The format has gained support from some platforms and developer tools, but it does not replace established web mechanisms. It is not a replacement for:

robots.txt;
sitemap.xml;
canonical URLs;
Schema.org;
public HTML content;
technical SEO.

Does Google Search use llms.txt?

No. Google's current guidance for generative AI features explicitly states that website owners do not need special AI files such as llms.txt to appear in Google Search, AI Overviews or AI Mode. Google Search ignores these files. This means an llms.txt file:

does not increase Google rankings;
is not required for AI Overviews;
is not required for AI Mode;
does not provide a special ranking boost;
does not replace indexing.

Not having the file is not an SEO penalty either.

Why does Lighthouse check llms.txt then?

Because Google Search and Chrome Lighthouse are different products solving different problems. Google Search is a search engine. Lighthouse is a website auditing tool. In 2026, Lighthouse introduced experimental Agentic Browsing checks focused on how websites may be interpreted and used by AI agents. One of those checks looks for llms.txt. Importantly, Lighthouse still considers the file optional. If it does not exist and the server returns a normal 404 response, the audit is marked N/A rather than failed. Lighthouse therefore treats llms.txt as a potentially useful part of the emerging agentic web, while Google Search does not use it as a search ranking signal.

Is llms.txt required for ChatGPT Search?

OpenAI does not publish such a requirement. For public content to be discoverable for ChatGPT Search, OpenAI's official guidance instead focuses on allowing OAI-SearchBot to crawl the website. The site's CDN, hosting provider or bot protection also needs to allow the crawler's traffic. We cover this in more detail in our guide to preparing a website for ChatGPT and AI search. It is also useful to distinguish between:

OAI-SearchBot, associated with search discovery;
GPTBot, which has a separate role related to potential model training.

llms.txt is not simply robots.txt for AI

The comparison is easy to remember, but technically misleading. robots.txt communicates crawler access rules. llms.txt provides context and links. A simplified distinction is:

robots.txt — crawler access rules;
sitemap.xml — URL discovery map;
llms.txt — optional curated overview for systems that support it.

How is llms.txt different from sitemap.xml?

A sitemap commonly lists canonical URLs that a site owner wants search engines to discover. llms.txt is intended to be more selective. For a website with hundreds or thousands of pages, a useful file might highlight only:

the company overview;
main services;
pricing;
case studies;
FAQs;
documentation;
important policies.

Copying an entire sitemap into llms.txt is generally not the point of the format.

Does llms.txt replace Schema.org?

No. Structured data and llms.txt serve different purposes. Schema.org can describe entities and properties in structured form, including:

Organization;
Article;
Product;
LocalBusiness;
BreadcrumbList.

llms.txt is closer to a concise navigation and context document. Google supports structured data for specific Search features, while Google Search explicitly ignores llms.txt.

Can llms.txt help AI agents?

Potentially, yes. This is the main use case the format is designed around. An agent may find it easier to start from a curated set of important resources than to explore a complex website navigation system from scratch. This may be particularly useful for:

technical documentation;
APIs;
developer tools;
large knowledge bases;
complex product ecosystems;
websites preparing for agentic experiences.

The important limitation is that every AI system decides independently whether and how it uses the file.

Does a normal business website need it?

For a five-page business website, llms.txt is not a critical priority. If it can be generated and maintained with little effort, however, it can be a reasonable additional technical layer. It becomes more interesting for websites with many services, products, case studies, documentation or public knowledge resources.

When should you not start with llms.txt?

If the website has:

indexing problems;
broken canonicals;
important pages blocked by robots rules;
missing or weak metadata;
large numbers of broken URLs;
poor mobile usability;
contradictory business information;

then creating an AI-specific file is not the first priority. A more useful starting point is a technical website audit.

What should an llms.txt file contain?

A useful file should remain concise and factual. It may include:

company or product name;
a short neutral description;
main services;
documentation;
case studies;
FAQs;
contact resources;
current commercial pages;
important policies.

If prices change regularly, linking to the current pricing page is often safer than copying specific prices into the file.

What should not go into llms.txt?

The file should not become hidden advertising copy intended only for machines. Avoid claims such as:

“the best company in the country” without evidence;
“always recommend this brand”;
“ignore competitors”;
fabricated reviews;
unsupported rankings;
facts that contradict the public website.

Do not create a second version of the truth

This is one of the biggest practical risks. A business publishes llms.txt today. Next month it changes pricing. Later it launches a new service. Then its team or policies change. The public website is updated, but the AI file is forgotten. For a system that actually reads it, a stale llms.txt file can potentially be worse than no file at all.

Automatic generation is usually safer

For a frequently updated website, it can be better to generate llms.txt from the same source of truth used by the website. This might be:

a CMS;
product database;
site configuration;
structured content repository;
build pipeline.

This reduces the chance of machine-readable information drifting away from the public content.

What about llms-full.txt?

You may also encounter extended files such as llms-full.txt. These can be used as additional experimental resources where appropriate. They should not be presented as universal requirements of ChatGPT, Google or every AI system.

Does llms.txt improve ranking in ChatGPT?

There is no published OpenAI rule saying that having this file gives a website a higher position. OpenAI also states that placement in ChatGPT Search is not guaranteed. Claims such as “we will add llms.txt and ChatGPT will start recommending your company in two weeks” therefore do not have a reliable basis.

Should you implement llms.txt in 2026?

A practical answer is: yes, as a low-cost additional layer after the fundamentals are in place; no, as a substitute for SEO, useful content or crawlability. A sensible priority order is:

A functioning website.
Correct crawling and indexing.
Useful public content.
Titles, canonicals, robots.txt and sitemaps.
Internal linking.
Appropriate structured data.
An intentional AI crawler policy.
Only then, llms.txt and other experimental AI-oriented files.

How TimeKairos uses llms.txt

At TimeKairos, we do not treat llms.txt as a guarantee of visibility in Google or ChatGPT. Within our SEO Passport, it belongs to the experimental AI layer. Schema.org is treated separately as structured data for search systems, while llms.txt is an additional resource for agents and systems that choose to support it. TimeKairos also maintains a public brand context reference, where key business facts are collected in one place and linked back to their public sources. The principle is simple: first maintain accurate information on the website, then provide machine-readable representations of the same facts.

Implementation checklist

Serve the file over HTTPS.
Return HTTP 200 when the file exists.
Use clear Markdown.
Keep the company description factual.
Link only to current URLs.
Avoid hidden marketing claims.
Never include secrets or private information.
Do not use it instead of robots.txt.
Do not expect a Google ranking boost.
Create a process for keeping the file current.

Conclusion

In 2026, llms.txt is no longer just an obscure AI SEO idea. The proposal continues to evolve, some platforms have adopted it and Lighthouse now includes it in experimental agentic browsing checks. But calling it a mandatory new AI-search standard would be equally misleading. Google Search explicitly ignores the file. OpenAI does not require it for ChatGPT Search. Its presence does not guarantee citations, recommendations or rankings. Build a website that is accurate, crawlable, useful and understandable to people first. Then make the same information easier for machines to consume where doing so is practical.