Free Download
Recent Posts
The web has always evolved alongside the tools people use to navigate it. First came search engines, and websites adapted with robots.txt and sitemaps. Now AI assistants are becoming a primary way people find information, and the web is adapting again. Enter the llms.txt file — a lightweight but potentially significant addition to your website that tells AI systems exactly what your site is about and where your best content lives.
This guide covers everything you need to know: what a llms.txt file is, how it works, what it should contain, detailed examples, and an honest look at what it can and cannot do for your website’s visibility in AI-powered search.
What Is a llms.txt File?
A llms.txt file is a plain-text Markdown document placed at the root directory of a website — accessible at yourwebsite.com/llms.txt — that provides structured, AI-readable information about your site’s content and purpose.
The standard was proposed in September 2024 by Jeremy Howard of Answer.AI and is documented at llmstxt.org. It gives large language models (LLMs) like ChatGPT, Claude, Gemini, and Perplexity a curated, easy-to-parse summary of your most important pages, rather than forcing them to scrape and interpret your full website from raw HTML.
The simplest way to think about it: llms.txt is to AI assistants what robots.txt is to search engine crawlers — a standardized file in a known location that communicates how your site should be understood. Unlike robots.txt, which tells crawlers what they may not access, llms.txt tells AI systems what they should pay attention to.
As of 2026, over 844,000 websites have implemented llms.txt, including major companies like Anthropic, Cloudflare, and Stripe. It remains a proposed standard — not yet formally adopted by any major AI provider — but adoption is growing rapidly as the industry moves toward AI-first content discovery.
The Problem llms.txt Was Designed to Solve
To understand why llms.txt matters, it helps to understand how AI assistants actually read your website — and how different that is from how a human reads it.
When someone asks ChatGPT or Perplexity a question that requires fetching information from your website, the AI doesn’t see the polished, visual version of your page. It sees the raw HTML source, which on most modern websites is cluttered with:
- Navigation menus and footers
- JavaScript files and tracking scripts
- Advertising code and cookie consent banners
- CSS stylesheets and layout wrappers
- Schema markup and meta tags
None of that is the content your visitors came to read. But AI models have to wade through all of it to reach your actual text — and they’re doing so under a significant constraint: context windows.
A context window is the maximum amount of text an AI model can process at one time. It’s finite. When an AI burns through its available context parsing navigation elements and JavaScript before it ever reaches your main content, the result is incomplete understanding, missed pages, and sometimes inaccurate information about your business.
This is the problem llms.txt addresses. By providing a clean, structured Markdown file in a predictable location, it gives AI systems a shortcut: a human-curated summary of your most valuable content without the noise.
How llms.txt Differs from robots.txt and Sitemaps
These three files are often mentioned together, but they serve very different purposes.
| File | Designed For | Purpose |
|---|---|---|
robots.txt | Search engine crawlers | Controls access — tells bots what they cannot crawl |
sitemap.xml | Search engines (Google, Bing) | Discovery — lists all pages for indexing |
llms.txt | AI language models | Understanding — explains what your site is and what matters most |
A sitemap lists every URL on your site with equal weight; it’s a comprehensive inventory. llms.txt is editorial — it’s a curated shortlist of your most important pages, written in plain language that an AI can immediately understand in context.
Importantly, llms.txt does not conflict with robots.txt. The two files serve different audiences and can coexist. Content blocked in robots.txt should generally not be linked in llms.txt, but otherwise they operate independently.
The llms.txt File Format and Structure
The llms.txt specification uses Markdown formatting, which is both human-readable and AI-friendly. The file must follow a specific order of sections, though only one section is technically required.
Required: H1 Title
Every llms.txt file must begin with an H1 heading containing the name of your project or website.
# Your Website Name
Recommended: Blockquote Summary
Immediately following the H1, a blockquote provides a concise summary of what your site is, who it serves, and what makes it useful. This is the most important descriptive element — AI systems use this as their primary context about you.
> A brief, factual description of your site in one to two sentences. > Include what you do, who you serve, and your main value proposition.
Optional: Additional Details
After the blockquote, you can include free-form Markdown content — paragraphs, bullet lists, anything except headings — that provides more background about your organization, products, or services.
H2 Sections: Categorized Link Lists
The bulk of a llms.txt file consists of H2-headed sections, each containing a list of important URLs formatted as Markdown links with optional descriptions.
## Section Name - [Page Title](https://yoursite.com/page): Brief description of what this page contains - [Another Page](https://yoursite.com/another): What a visitor (or AI) will find here
The “Optional” Section
The specification includes a special H2 section called ## Optional. Links placed under this heading signal to AI systems that this content can be skipped when context is limited. Use it for supplementary resources that are useful but not critical.
## Optional - [Case Studies](https://yoursite.com/case-studies): Real-world examples and results - [Blog Archive](https://yoursite.com/blog): Historical articles and industry commentary
Detailed llms.txt Examples
Example 1: Local Service Business (Plumbing Company)
This example shows a straightforward llms.txt for a local business with a simple site structure.
# Riverside Plumbing & Drain > Riverside Plumbing & Drain is a licensed plumbing contractor serving the greater Sacramento, California area since 1994. We specialize in residential and commercial plumbing repair, drain cleaning, water heater installation, and emergency service available 24/7. Family-owned and operated. Licensed, bonded, and insured in California (License #C-36 452718). We serve Sacramento, Elk Grove, Roseville, Folsom, and surrounding communities within a 40-mile radius. ## Services - [Drain Cleaning & Unclogging](https://riversideplumbing.com/drain-cleaning): Hydro-jetting, snaking, and camera inspection for residential and commercial drains - [Water Heater Installation & Repair](https://riversideplumbing.com/water-heaters): Traditional tank, tankless, and heat pump water heater services - [Emergency Plumbing](https://riversideplumbing.com/emergency): 24/7 emergency service with same-day response for urgent plumbing failures - [Sewer Line Repair](https://riversideplumbing.com/sewer-line): Trenchless pipe lining and traditional excavation repair - [Bathroom & Kitchen Plumbing](https://riversideplumbing.com/fixture-installation): Faucet, toilet, and fixture installation and replacement ## Company Information - [About Us](https://riversideplumbing.com/about): Company history, licensing, and service area - [Service Area](https://riversideplumbing.com/service-area): Full list of cities and ZIP codes we serve - [Pricing & Estimates](https://riversideplumbing.com/pricing): Flat-rate pricing and how to request a free estimate - [Contact](https://riversideplumbing.com/contact): Phone number, address, and online contact form ## Optional - [Reviews & Testimonials](https://riversideplumbing.com/reviews): Customer reviews from Google and Yelp - [Plumbing Tips Blog](https://riversideplumbing.com/blog): DIY maintenance advice and home plumbing guides - [FAQ](https://riversideplumbing.com/faq): Common questions about services, pricing, and scheduling
Example 2: SaaS Product / Software Company
# TaskFlow — Project Management for Remote Teams > TaskFlow is a cloud-based project management platform built for distributed teams. It combines task tracking, time logging, client billing, and team communication into a single workspace, with native integrations for Slack, GitHub, and Google Workspace. Available on web, iOS, and Android. Plans start at $9/user/month. SOC 2 Type II certified. Supports teams from 5 to 5,000+ users. Data hosted in US, EU, and AU regions. ## Product - [Features Overview](https://taskflow.io/features): Complete list of platform capabilities including task management, time tracking, Gantt charts, and reporting - [Pricing](https://taskflow.io/pricing): Starter, Professional, and Enterprise plan details with feature comparison - [Integrations](https://taskflow.io/integrations): Full list of 80+ native integrations including Slack, GitHub, Zapier, and Salesforce - [Security & Compliance](https://taskflow.io/security): SOC 2 certification, data residency options, SSO, and enterprise security controls - [What's New](https://taskflow.io/changelog): Product updates and release notes ## Documentation - [Getting Started Guide](https://docs.taskflow.io/getting-started): How to set up your workspace and invite your team - [API Reference](https://docs.taskflow.io/api): REST API documentation for developers building integrations - [Importing Data](https://docs.taskflow.io/import): Guides for migrating from Asana, Trello, Monday.com, and Jira - [Admin Controls](https://docs.taskflow.io/admin): User permissions, billing management, and organization settings ## Support - [Help Center](https://help.taskflow.io): Searchable knowledge base with tutorials and troubleshooting guides - [Status Page](https://status.taskflow.io): Real-time platform uptime and incident history - [Contact Support](https://taskflow.io/support): Submit a ticket or start a live chat ## Optional - [Customer Stories](https://taskflow.io/customers): Case studies from teams using TaskFlow - [Blog](https://taskflow.io/blog): Articles on remote work, productivity, and project management - [Webinars & Events](https://taskflow.io/events): Live training sessions and recorded webinars - [Affiliate Program](https://taskflow.io/affiliates): Partner program details and commission structure
Example 3: E-Commerce Store
# Northwoods Outdoor Gear > Northwoods Outdoor Gear is an online retailer specializing in camping, hiking, backpacking, and survival equipment. We carry 4,000+ products from top brands including MSR, Black Diamond, Osprey, Sea to Summit, and REI Co-op. Free shipping on orders over $75. Based in Duluth, Minnesota. All products are in-stock at our Duluth warehouse unless noted. We offer a 60-day return policy on unused gear. Customer service available Monday–Friday 9am–5pm CST. ## Shop by Category - [Tents & Shelters](https://northwoodsoutdoor.com/tents): 3-season, 4-season, and ultralight backpacking tents - [Sleeping Bags & Pads](https://northwoodsoutdoor.com/sleep): Down and synthetic sleeping bags rated from +20°F to -40°F - [Backpacks & Bags](https://northwoodsoutdoor.com/packs): Daypacks, overnight packs, and expedition packs up to 90L - [Cooking & Water](https://northwoodsoutdoor.com/cooking): Camp stoves, cookware, water filters, and purification tablets - [Clothing & Footwear](https://northwoodsoutdoor.com/clothing): Layering systems, rain gear, and hiking boots - [Navigation & Safety](https://northwoodsoutdoor.com/navigation): GPS devices, compasses, first aid kits, and emergency gear ## Customer Service - [Shipping Policy](https://northwoodsoutdoor.com/shipping): Shipping rates, estimated delivery times, and carrier information - [Returns & Exchanges](https://northwoodsoutdoor.com/returns): 60-day return policy details and how to initiate a return - [Size & Fit Guides](https://northwoodsoutdoor.com/sizing): Pack fit guides, boot sizing, and layering recommendations - [Contact Us](https://northwoodsoutdoor.com/contact): Phone, email, and live chat options ## Optional - [Gear Reviews & Guides](https://northwoodsoutdoor.com/blog): Buying guides, gear comparisons, and trip planning resources - [Brand Directory](https://northwoodsoutdoor.com/brands): Full list of brands we carry - [Gift Cards](https://northwoodsoutdoor.com/gift-cards): Digital gift cards in any amount
Example 4: Technical Documentation / API Site
This format — closer to how companies like Anthropic and Stripe structure their llms.txt — is well-suited for developer-focused documentation.
# Acme API > Acme provides a REST and GraphQL API for payment processing, subscription billing, and financial data management. SDKs are available for Node.js, Python, Ruby, PHP, Go, and Java. Used by 12,000+ businesses processing over $2 billion annually. API version: v3.1. Base URL: https://api.acme.io/v3. Authentication via API keys and OAuth 2.0. Webhooks supported for all major events. Rate limit: 1,000 requests/minute on standard plans. ## Core Documentation - [API Overview](https://docs.acme.io/overview): Architecture, authentication, and request format - [Authentication](https://docs.acme.io/auth): API key setup, OAuth flows, and security best practices - [Payments API](https://docs.acme.io/payments): One-time charges, payment methods, and refunds - [Subscriptions API](https://docs.acme.io/subscriptions): Creating plans, managing billing cycles, and proration - [Webhooks](https://docs.acme.io/webhooks): Event types, payload format, and signature verification - [Error Reference](https://docs.acme.io/errors): Complete list of error codes and how to handle them ## SDK Guides - [Node.js SDK](https://docs.acme.io/sdk/node): Installation, configuration, and usage examples - [Python SDK](https://docs.acme.io/sdk/python): pip installation and quick-start examples - [PHP SDK](https://docs.acme.io/sdk/php): Composer installation and WordPress/Laravel integration notes ## Optional - [Migration Guide (v2 → v3)](https://docs.acme.io/migration): Breaking changes and upgrade path - [Postman Collection](https://docs.acme.io/postman): Downloadable collection for testing endpoints - [Changelog](https://docs.acme.io/changelog): Full version history and deprecation notices
llms-full.txt: The Companion File
The llms.txt specification also defines a companion file: llms-full.txt.
Where llms.txt is an index — a curated navigation guide pointing to important URLs — llms-full.txt is a comprehensive single-file document that contains the actual content of your key pages, all concatenated together in Markdown format.
The practical difference:
- llms.txt tells an AI where to find your content
- llms-full.txt gives an AI the content itself, without requiring it to follow links
llms-full.txt is most valuable for technical documentation sites, API reference guides, and knowledge bases where an AI assistant might need deep context about your product in a single fetch. For most marketing sites and blogs, a well-structured llms.txt is sufficient.
The naming convention matters: the specification requires llms.txt (lowercase, exact spelling — not llm.txt or LLMs.txt) to ensure cross-platform compatibility and reliable AI discovery.
llms.txt and SEO: What It Can and Cannot Do
This is where honest assessment matters most, because there is significant hype around llms.txt and its effects — and not all of it is accurate.
What llms.txt Cannot Do
It will not improve your rankings in Google Search. In July 2025, Google’s Gary Illyes confirmed explicitly that Google does not support llms.txt and has no plans to do so. John Mueller compared it to the discredited keywords meta tag. Traditional Google SEO is unaffected by this file.
It is not a guaranteed signal for any major AI platform. As of mid-2026, no major LLM provider — OpenAI, Anthropic, Google, Meta — has publicly committed to reading llms.txt as a formal signal in their production systems. Analysis of AI bot traffic across hundreds of millions of visits shows that requests specifically targeting /llms.txt represent a statistically small fraction of total AI crawler activity.
What llms.txt Can Do
Reduce AI hallucinations about your brand. When an AI model has a clean, structured, authoritative description of your business from your own llms.txt file, it has less reason to invent or misrepresent details. This is probably the most concrete and immediate benefit.
Improve accuracy when AI systems do access your content. If a crawler or AI system does fetch your llms.txt, it gets accurate, curated information rather than having to interpret noisy HTML. The file is a low-risk way to make the most of any AI traffic you do receive.
Provide an edge as the standard matures. Adoption of web standards often happens before formal support is announced. robots.txt and structured data both had adoption periods before major platforms fully embraced them. Implementing llms.txt now positions you ahead of the curve at essentially no cost.
Improve AI search visibility in emerging platforms. AI-powered search tools like Perplexity, SearchGPT, and others are evolving rapidly. As these platforms formalize how they discover and rank content, having a well-structured llms.txt could become a meaningful signal.
Benefit your internal AI workflows. Even if no external AI ever reads your llms.txt, you can use it yourself — upload it to a Claude Project, include it in your Cursor or Copilot context, or reference it when working with AI on content strategy and audits. The process of writing it also forces a useful exercise: clearly articulating what content on your site actually matters most.
The Honest Position
Implementing llms.txt is a low-effort, no-downside action. The file takes 30–60 minutes to create, has no negative SEO implications, and positions your site well if and when AI platforms formalize support. The risk of not having one is low today; the risk of still not having one in two years could be higher. Think of it as infrastructure you’re laying ahead of demand.
How to Create a llms.txt File
Step 1: Create the File
Open any text editor and create a new file named llms.txt (all lowercase). Save it with UTF-8 encoding.
Step 2: Write Your H1 Title
# Your Website or Business Name
Step 3: Write Your Blockquote Summary
This is the most important piece of writing in the file. Be factual, specific, and concise.
> [Your business name] is a [what you are] that [what you do] for [who you serve]. > [One additional sentence about your location, scale, or key differentiator if relevant.]
Step 4: Add Optional Context (if useful)
A brief paragraph about your founding year, certifications, geographic coverage, business model, or other facts that help AI understand your organization’s scope and credibility.
Step 5: Build Your H2 Sections
Group your most important pages into logical categories. Aim for pages that, if an AI read them, would give it an accurate and complete picture of your site’s value. Good candidates include:
- Your main service or product pages
- Pricing and plans
- Contact and location information
- Key documentation or guides
- About and credentials pages
Step 6: Add an Optional Section
Move secondary content — blog archives, case studies, press pages — into the ## Optional section.
Step 7: Place the File at Your Root
Upload llms.txt to the root directory of your website so it’s accessible at https://yourwebsite.com/llms.txt. Test it by visiting that URL in your browser.
WordPress Users
Several SEO plugins now generate llms.txt automatically. Yoast SEO added llms.txt support, and dedicated plugins are available in the WordPress repository. These are useful starting points, but review the generated output and customize it — automated generation often produces generic descriptions that miss the nuance a manually written file provides.
Common Mistakes to Avoid
Skipping the blockquote. Without the blockquote, AI systems lose the structured cue that signals “this is the summary.” They fall back to inferring context from individual link descriptions, which is slower and less reliable.
Including broken links. Any link in your llms.txt that returns a 404 error undermines the file’s usefulness. Treat this file like production infrastructure — audit it periodically and update it when pages are removed or URLs change.
Adding nested headings inside H2 sections. The structure is: one H1, one optional blockquote, optional free-form content, then H2 sections with link lists. H3 or H4 headings inside an H2 section are outside the spec and may cause parsing issues.
Keyword stuffing link descriptions. Descriptions should be clear, factual, and written for AI comprehension — not padded with target keywords the way meta descriptions once were. Treat this file as a briefing document, not marketing copy.
Listing every page on your site. llms.txt is a curated shortlist, not a sitemap. If you’re including more than 30–40 links, you’re probably over-indexing. Focus on pages that provide meaningful context about what you do and who you are.
Pointing to content behind a login or paywall. AI crawlers cannot access authenticated content. Any URLs in llms.txt must be publicly accessible.
Who Benefits Most from llms.txt?
While any website can implement llms.txt, some are positioned to benefit more than others in the near term.
Developer tools and SaaS products see the clearest benefit. Technical users regularly feed documentation URLs to AI coding assistants. A well-structured llms.txt or llms-full.txt means those interactions produce more accurate, complete responses. It was developer-focused companies — Stripe, Anthropic, Supabase, Cursor — who were first to adopt the standard.
Local service businesses have a strong use case for accuracy. AI assistants are increasingly asked “who’s the best plumber in [city]” or “find me an electrician near me.” Having a clearly structured llms.txt with accurate service area information, licensing details, and service descriptions reduces the chance of AI systems representing your business incorrectly.
E-commerce sites benefit from ensuring AI has accurate product category, shipping policy, and brand information — especially as conversational commerce grows.
Content-heavy sites like news publishers, educational platforms, and knowledge bases can use llms.txt to direct AI toward their highest-quality, most authoritative content rather than letting crawlers find whatever comes up first.
Businesses sensitive to misrepresentation — medical practices, legal firms, financial services — have a practical interest in providing authoritative, accurate descriptions of what they do and don’t offer, reducing the likelihood of AI systems overstating their scope.
The Bottom Line
A llms.txt file is a small investment with a potentially significant long-term payoff. It won’t move your Google rankings today, and the AI platforms that matter haven’t formally committed to reading it. Those are honest limitations worth knowing.
What it does do: give AI systems a clean, structured, authoritative description of your website directly from the source. As AI-powered search and conversational interfaces continue to grow — and they are growing rapidly, with AI assistants now handling a substantial fraction of information queries that once went to Google — the websites that have made themselves legible to these systems will have an advantage over those that haven’t.
The file takes an hour to write well. There is no downside to having one. The question isn’t really whether it’s worth doing — it’s whether you’d rather do it now or catch up later.
This article was last updated June 2026. The llms.txt standard is actively evolving. For the most current specification, refer to llmstxt.org.
