What is llms.txt and why does it matter?
llms.txt is a proposed plain-text site summary for machines. It is cheap to publish and Chrome Lighthouse now checks its format for agentic browsing, but Google says it is not needed for Search and no major answer engine has confirmed that it uses the file for retrieval or citation. Here is what the evidence supports in 2026.
Video transcript
You've heard of robots.txt. Meet the file built for AI: llms.txt. It tells ChatGPT, Claude and Perplexity what your site is and what to read. Miss it, and they guess. First, it's plain text. One simple file at the root of your domain. Second, you're in control. You choose what AI sees, and what to prioritise. Third, it's a rising standard. Early adopters are already getting cited more. Comment 'LLMS' for the template, free on SearchScore.
Key takeaway: llms.txt is a low-cost, optional machine-readable site summary. Publish one if you want to support emerging agent workflows, but do not treat it as an AI-search ranking or citation tactic. Google says it does not affect Search visibility, and SearchScore has found no evidence that the major answering engines use it at retrieval time.
The problem llms.txt solves
When an AI language model crawls your website, it faces a challenge. Your website was designed for humans - with visual design, navigation, images and layout that help people find what they need. An AI model reading your HTML has to figure out what is important, what is marketing fluff, what is navigation boilerplate, and what is the core content it should actually process.
llms.txt attempts to solve this by giving compatible tools a clean, structured summary: who you are, what you do and which pages matter. The important qualifier is compatibility: publishing the file does not make an answering engine read it, and it is not a substitute for crawlable HTML, ordinary search indexing or structured content.
What does an llms.txt file look like?
The format is simple Markdown. Here is an example for a hypothetical digital agency:
# Acme Digital Agency
> We help B2B SaaS companies grow through SEO, paid search and content marketing.
## About
Acme Digital Agency is a full-service digital marketing agency founded in 2018.
We work with Series A to Series C SaaS companies across the UK and US.
Our team of 24 specialists covers SEO, PPC, content, analytics and CRO.
## Key pages
- [Services](https://acmedigital.com/services/): Our core service offerings
- [Case Studies](https://acmedigital.com/case-studies/): Client results and outcomes
- [Blog](https://acmedigital.com/blog/): Guides and industry insights
- [About](https://acmedigital.com/about/): Team and company background
## Key topics
SEO strategy, B2B lead generation, SaaS marketing, content marketing, conversion rate optimisation.
## Contact
[email protected] | London, UK
llms.txt vs robots.txt: what is the difference?
| File | Purpose | Audience |
|---|---|---|
robots.txt |
Controls which bots can crawl which pages | All web crawlers |
llms.txt |
Optional machine-readable summary of site content and purpose | Compatible agents/tools; not confirmed as an AI-search retrieval input |
sitemap.xml |
Lists all pages for crawling and indexing | Search engine crawlers |
You need all three. They serve different purposes and are not substitutes for each other.
Google Cloud added a further format in June 2026, the Open Knowledge Format (OKF). It is well designed, but the consumer AI engines that decide whether ChatGPT or Perplexity recommend you do not read it, which is why our research treats OKF as not urgent for AI visibility.
llms-full.txt: the extended version
Alongside llms.txt, some websites publish an llms-full.txt file containing longer extracts from important pages. This can be useful to tools that deliberately support the convention, but there is no evidence that publishing it causes ChatGPT, Gemini, Claude or Perplexity to ingest or cite the content.
If you have key pages - a detailed about page, your flagship guide, your most important service description - consider including their full text in llms-full.txt.
What our data says: In SearchScore’s August 2026 study of 102,873 domains, sites publishing an llms.txt scored 5.36 points higher on average, but 1.37 points came from the scorer rewarding the file itself. A follow-up crawl also found that 41.8% of sampled URLs recorded as present actually returned HTML rather than a genuine file. The evidence supports cheap housekeeping and agent readiness, not a claim that llms.txt improves citation rankings. See the study and the compliance crawl.
How to create and deploy your llms.txt
1. Create a plain text file named llms.txt
2. Write a one-line description of your website after a > character
3. Add an About section describing what you do and who you serve
4. List your 5 to 10 most important pages with their URLs and one-line descriptions
5. Add a Key Topics section naming the subjects your site covers
6. Upload the file to the root directory of your website (so it is accessible at yoursite.com/llms.txt)
The whole process takes under 30 minutes for most websites. It is one of the simplest GEO improvements available.
Back to pillar
- What is GEO? The Complete Guide →
S
Ronnie Huss
GEO Research & Analysis
The SearchScore editorial team researches and writes about generative engine optimisation, AI search visibility and the signals that determine whether your website gets cited by ChatGPT, Perplexity and Google AI Overviews.
Sources & Further Reading
- llmstxt.org – The llms.txt standard specification
- OpenAI – OpenAI crawlers: OAI-SearchBot, GPTBot and ChatGPT-User
- SearchScore SAVI Report, Q3 2026 (Vol. 3, 1,000,000+ sites analysed)
Check your AI visibility
Enter your URL at SearchScore for a free AI visibility Score. See how ChatGPT, Perplexity and Google AI see your site - and exactly what to fix.
Frequently asked questions
What is llms.txt?
llms.txt is a plain text file placed in the root of a website (e.g. yoursite.com/llms.txt) that provides structured guidance for AI language models about the site's purpose, key content and structure. It is modelled on robots.txt but designed for LLMs rather than traditional search crawlers.
Is llms.txt an official standard?
llms.txt was proposed by Answer.AI in 2024 and is not a W3C standard. It does now have a published pass condition: Chrome's Lighthouse added an Agentic Browsing category in August 2026 that opens the file and checks it has a top-level heading and at least one markdown link. That is a real test rather than an adoption claim. It is not evidence that any AI engine reads the file, and no provider has confirmed that it does.
Do I need technical knowledge to create an llms.txt file?
No. llms.txt is plain text written in simple Markdown format. Anyone who can create and upload a text file to a website can create an llms.txt. The format is intentionally simple.