Is llms.txt Worth Adding to Your Site? (October 2026)

For most small business clients, no. llms.txt is a file that tells AI systems what to read, but in October 2026 the evidence is clear: ChatGPT, Claude, Gemini, and Perplexity rarely request it, no study has found a citation advantage, and it solves a problem your clients do not have yet.

What llms.txt proposes

The llms.txt file is a standardized markdown document hosted at a website's root path (tryprofound.com) (such as yoursite.com/llms.txt). The file uses Markdown to structure information rather than a classic structured format such as XML, because these files are expected to be read by language models and agents.

The file contains a site name as an H1 heading, a blockquote description, and then sections with H2 headings that organize links to important pages. AI systems typically fetch llms.txt at inference time when users explicitly request information about a business, and the file is designed for quick parsing and context extraction (ai-visibility.org.uk).

A companion file, llms-full.txt, can include the full text of documentation or content in a single file. The idea is to give AI systems clean, readable content instead of making them parse HTML with navigation, ads, and tracking code.

The official specification is open and was proposed by Jeremy Howard in 2024. Documentation platforms like Mintlify, Fern, GitBook, and ReadMe now generate llms.txt files automatically (mintlify.com).

Which assistants actually honour it today

None of them do, at scale.

Google publicly confirmed Google Search doesn't read or act on llms.txt, and Gary Illyes from Google added Google doesn't support the standard and isn't planning to (dataforaisearch.com). OpenAI, Anthropic and Google have not documented it as a crawl, citation or ranking input (quora.com).

A study by Ahrefs, published in June 2026 and drawing on server-log data from 137,000 domains, found that 97% of llms.txt files received zero requests in May 2026 (ppc.land). Even where AI retrieval bots (those that answer user queries inside AI search products) were active, they accounted for just 1.1% of requests to llms.txt files.

A separate analysis covering over 500M LLM bot traffic events found that AI search crawlers are almost never fetching /llms.txt, and GPTBot, ClaudeBot, PerplexityBot, OAI-SearchBot, and Google-Extended overwhelmingly skip the file and crawl HTML directly.

The one working use case is AI coding tools. llms.txt does have a real, working use case: AI coding tools and agentic browsers, and Cursor, Windsurf, GitHub Copilot, Claude Code, Cline, and similar tools do look for /llms.txt and /llms-full.txt when they're pointed at a documentation site. That matters if you sell to developers who use those tools, which most small business clients do not.

Does it change anything yet?

No. Both statistical analysis and machine learning showed no effect of LLMs.txt on how often a domain is cited by LLMs, according to SE Ranking's 300,000-domain study in 2025. A large study analyzing around 300,000 domains found no meaningful correlation between having an llms.txt file and how often a site gets cited in AI generated answers, and removing llms.txt as a factor from their prediction model actually made the model more accurate.

5.61% of the top 10,000 websites have a valid llms.txt as of June 2026. Adoption has grown, but traffic has not. Originality.ai counted 38,980 adopting sites across 3M+ websites by May 2026, yet SEO audit tools (not AI assistants) dominate actual llms.txt request logs.

The gap between hype and reality is wide enough that it is worth naming plainly: agencies have started selling llms.txt implementation services, but the thing those services promise to improve does not exist yet.

When would you actually recommend it?

In three cases, and only three.

First, your client sells developer tools or hosts API documentation, and their users work in Cursor or GitHub Copilot every day. Stripe, Vercel, Cloudflare, Anthropic, Coinbase, Pinecone, Cursor, and most modern API products ship llms.txt because their users are building with AI coding assistants right now. For that audience the file does real work today.

Second, your client is preparing for a future where autonomous agents browse the web and they are willing to invest time without a current return. The file costs a few hours to create and maintain, and the downside is near zero if it is consistent with robots.txt. That is a reasonable bet for some businesses.

Third, never. For a local plumber, a chiropractor, or a boutique retail shop, llms.txt solves nothing and the time is better spent getting cited the way citations actually happen today: by being a trusted, well-linked source on the topics customers ask about. You can check whether your clients appear when someone asks Claude, ChatGPT, Gemini, or Perplexity who to hire using the free AI-visibility check (/check), and if they do not, the reason is not a missing text file.

If a client asks about it, the honest answer is that it might matter in twelve months but it does not matter now. You can create one in an afternoon with a generator or by hand, but do not sell it as a visibility strategy and do not charge for it as if it were one. The evidence is in, and the evidence says it changes nothing yet.

For tracking what actually drives AI citations across Claude, ChatGPT, Gemini, and Perplexity, ActiScore's pricing (/pricing) starts at $79 per month and includes side-by-side competitor audits on the same rubric those assistants use. If you want to see how your client's technical foundation compares before committing time to speculative standards, the guide (/guide) walks through the eight categories measured in every scan.