For most small business websites, no. Google says plainly that you don't need llms.txt or any other "AI text file" to appear in Google Search, AI Overviews or AI Mode, and we found no statement from OpenAI or Perplexity that their search tools use it to choose sources. Google says maintaining one neither helps nor harms visibility in Google Search. It can be a useful optional extra for documentation that AI agents may read, provided you keep it public-safe and accurate. It isn't a fix for being missing from AI answers.

What llms.txt is

llms.txt is a proposed standard, published by Jeremy Howard in September 2024, for a plain-text Markdown file at the site root or a relevant subdirectory (for example example.com/llms.txt). It gives AI tools a short, curated summary of the site: an H1 title, optionally followed by a summary and grouped links to useful pages, ideally with clean Markdown versions of those pages (llmstxt.org, checked 2026-10-19).

The idea is that AI agents could read a tidy summary instead of working through menus, pop-ups and scripts. The proposal is mainly aimed at use "at inference time", meaning when an AI tool is answering a question, rather than for training.

A minimal example for a hypothetical business:

# [Business name]

> [Business name] is a commercial cleaning company serving offices in Singapore.

## Services
- [Office cleaning](https://example.com/services/office-cleaning/): daily and weekly contracts
- [Post-renovation cleaning](https://example.com/services/post-renovation/)

## Company
- [Contact](https://example.com/contact/): phone, address and hours

llms.txt is not robots.txt

The names are similar, but the jobs are different.

  • robots.txt gives compliant automated crawlers instructions about which URLs they may crawl. It is not an access-security mechanism, and user-requested fetchers may handle it differently (Google robots.txt guidance; OpenAI crawlers; Perplexity crawlers).
  • llms.txt is a suggested summary. It doesn't allow or block anything.

To manage automated crawling, use the provider's documented robots.txt rules and other content controls. These do not govern every user-requested visit or every use of content.

What Google says

Google's position is the clearest of any major search provider.

  • In its guide to optimising for generative AI features, Google says: "You don't need to create new machine readable files, AI text files, markup, or Markdown to appear in Google Search." The guide adds that creating such files for other services won't hurt your visibility in Search (Google Search Central, last updated 2026-07-10, checked 2026-10-19).
  • Its page on AI features says there are "no additional requirements to appear in AI Overviews or AI Mode, nor other special optimizations necessary". A page needs to be indexed and eligible to show with a snippet (Google Search Central, checked 2026-10-19).

Google also offers Settings > Search generative AI in Search Console to control inclusion in AI Overviews, AI Mode and generative AI features in Discover. This is separate from AI-training controls.

Why Lighthouse now checks for llms.txt

This is where the confusion comes from. In May 2026, Google's Lighthouse tool (built into Chrome's developer tools and used by many site-audit services) added an "Agentic browsing" section that includes an llms.txt audit (Lighthouse 13.3.0 release notes, 7 May 2026; Chrome for Developers: llms.txt audit, checked 2026-10-19).

Look closely at what it checks:

  • If the request fails or the server returns an HTTP 5xx error when Lighthouse asks for /llms.txt, the audit fails.
  • If your server returns a normal "not found" (404), the audit is marked "Not Applicable", because, in Lighthouse's own words, "providing the file is optional at the moment".

If a file is returned, Lighthouse also checks for an H1 heading, a Markdown link and at least 50 characters. These checks are present in the released Lighthouse 13.5.0 audit, checked on 19 October 2026.

Lighthouse's documentation does recommend creating the file. But this is about AI agents that browse and act on websites, not about how Google Search ranks pages or chooses sources for AI Overviews.

The practical lesson: if you don't have an llms.txt file, make sure asking for one returns a clean 404, not a server error and not your home page with a "200 OK" status. That's good housekeeping either way.

What ChatGPT, Perplexity and Copilot say

We checked each provider's documentation for site owners on 2026-10-19:

  • OpenAI documents OAI-SearchBot as a crawler used for ChatGPT search and provides robots.txt controls for it. Its crawler guidance does not identify llms.txt as a requirement for search inclusion or citation selection (OpenAI: Overview of OpenAI crawlers).
  • Perplexity says PerplexityBot surfaces and links websites in its results and recommends allowing it in robots.txt. Its crawler guidance does not identify llms.txt as a requirement for search inclusion or citation selection (Perplexity crawlers).
  • Microsoft Bing's guidance on being cited in Copilot focuses on clear, well-structured, up-to-date content and IndexNow, not llms.txt (Bing Webmaster Blog, 10 February 2026).

AI companies including OpenAI and Anthropic publish llms.txt files for their own developer documentation. Publishing a file for your own docs isn't the same as saying your search product reads other sites' files to choose citations. We found no such statement from any of them.

Third-party write-ups disagree about whether AI crawlers request llms.txt at all, and most are based on a single site's server logs. We haven't relied on them here.

When an llms.txt file is worth adding

Consider adding one if:

  • You publish documentation, a knowledge base or a product catalogue that developers or AI agents may want to read directly. This is where the format was designed to help.
  • Your website platform or SEO plugin can generate it automatically, so the maintenance effort may be small. Check what it produces before switching it on.
  • You want to tidy up a Lighthouse or site-audit warning caused by a server error on /llms.txt. In that case, either add a simple file or fix the error so a 404 is returned.

It's not worth much time if:

  • You run a typical small business site with a home page, a handful of service pages and a contact page. A search engine or AI tool can already read those pages directly.
  • Someone is charging you to add it as a way to "get into ChatGPT" or "rank in AI Overviews". Nothing in the official documentation supports that claim.

If you do add one, keep it simple and accurate

  • Put it at your site root or a relevant subdirectory and follow the structure on llmstxt.org: an H1 title, optionally followed by a summary and grouped links. The proposal requires the H1; Lighthouse applies the additional checks described above.
  • Link only to live pages you want people to find: services, prices or price ranges if you publish them, contact details, key guides.
  • Don't include anything you wouldn't put on your website. The file is public.
  • Update it when you add or remove important pages. An out-of-date summary is worse than none.

What helps more than llms.txt

Our guide to what AI citations are explains the wider picture. If your goal is to appear in AI answers, the things that matter are the same things the providers' own documentation points to:

  • Let the right crawlers reach your site (Googlebot, Bingbot, OAI-SearchBot, PerplexityBot) in robots.txt.
  • Check whether your important pages are indexed in Google and Bing, and address any eligibility problems. Indexing is not guaranteed.
  • Answer customers' questions clearly near the top of each page, in sections that stand on their own.
  • Back up facts with sources, and keep business details consistent across your website and listings.

None of this guarantees a citation, and llms.txt doesn't either. Which sources an AI tool picks for a given question is outside any site owner's control.