Google Search and Chrome documentation now point in different directions on llms.txt, depending on whether the goal is search visibility or browser agent readiness.
Google Search recently released a new optimization guide which lists llms.txt as one of the tactics you don’t need for generative AI features. The guide bundles it with content segmentation, AI-specific rewriting, and special schema.
A few days earlier, Google’s Lighthouse tool had delivered version 13.3, which added a new Agentic Navigation Category. The update includes a audit llms.txt which checks if a site provides the file and reports server errors when retrieving it.
Lighthouse’s documentation describes llms.txt as a way to provide “a machine-readable summary of a website’s content, designed specifically for LLMs and AI agents.” He adds that without the file, “agents can spend more time exploring the site to understand its high-level structure and main content.”
What Google Search Says
Google’s research team has maintained for over a year that llms.txt is not a Google initiative or something Google plans to adopt.
John Mueller compared llms.txt to keyword meta tagnoting that no AI services used it and bots did not request the file. He called for creating separate Markdown pages for bots “a stupid idea.»
HAS Search Central Live Deep Dive Asia PacificGary Illyes and Amir Taboul confirmed that Google is not pursuing llms.txt.
Google optimization guide explicitly states that llms.txt should be ignored, providing the most recent direct statement from the research team.
What Chrome Flagship Does Now
Lighthouse 13.3 ships with the Agentic Navigation category by default, checking WebMCP integration, agent accessibility, layout stability and llms.txt.
The llms.txt audit only marks sites as “Not Applicable” if they return a 404; errors flag the audit. Lighthouse documents describe llms.txt as an “emerging convention” on llmstxt.org, advising site owners to create it and place it in their root directory.
This category is separate from SEO audits and indicates that llms.txt helps browser-based agents understand site structure, not improve search rankings or AI citations.
Google has been here before
Internal Google teams have already sent mixed signals about llms.txt.
In December, Lidia Infante spotted an llms.txt file in Google’s Search Central developer documentation. Mueller responded on Bluesky with “hmmn :-/” and did not elaborate further.
Dave Smart note that the file appeared on several Google developer properties, including Developer.chrome.com and web.dev. The model suggested an internal update to the CMS platform that automatically deploys llms.txt files, not a decision by the research team.
The Search Central file was deleted within hours, but files for other Google properties remained.
Why it matters
Google’s response on llms.txt varies depending on the use case.
For Google Search, llms.txt is not required for AI previews, AI mode, or other generative AI search features.
For browser-based agents, Lighthouse considers llms.txt optional in an experimental machine interaction category.
The guidance is spread across different Google developer sites, which can result in conflicting instructions when comparing Lighthouse or its llms.txt documentation with Google’s search docs.
Looking to the future
Google has not commented on the documentation gap between the two product teams.
For many sites, creating a basic llms.txt file is simple, but maintaining it is questionable, given that Google Search indicates that it is not necessary for AI Search visibility.
Featured Image: Stock-Asso/Shutterstock
