llms.txt & ai.txt Generator — make your site AI-friendly
Generate llms.txt, llms-full.txt, or ai.txt for your domain in under a minute. These files tell ChatGPT, Claude, Gemini and Perplexity what your site is about and how to reference it — the new robots.txt for AI search.
Each line in Pages: Title | URL | Optional description
What llms.txt unlocks in 2026
Generative Engine Optimization (GEO) is the new SEO. These files are how you tell answer engines who you are and what to cite.
Get cited in ChatGPT answers
When ChatGPT browses to draft a reply, an llms.txt helps it pick the canonical pages on your site to quote and link back to.
Surface in Perplexity citations
Perplexity loves clean, structured source lists. A clear llms.txt makes your most authoritative pages discoverable.
Help Claude understand your docs
For dev tools and APIs, llms-full.txt lets Claude ingest full content without crawling JavaScript-heavy pages.
AI Overviews ranking
Google's AI Overviews increasingly pull from sites that publish machine-readable summaries alongside human pages.
Future-proof your SEO
Robots.txt was the 1994 protocol for search engines. llms.txt is its 2026 equivalent for the answer-engine era.
Documentation sites
Stripe, Cloudflare, Anthropic and Vercel all publish llms.txt for their developer docs — you should too.
How to publish llms.txt in 4 steps
Fill the form
Enter your site name, URL, and short description. Add the sections you want to expose — Docs, Blog, Pricing, Changelog, whatever matters most.
Pick a mode
Use llms.txt for a curated index, llms-full.txt when you want to embed full page bodies, or ai.txt for Spawning.ai-style opt-out directives.
Copy or download
Hit Copy or Download in the output pane. The result is a plain UTF-8 text file ready for production.
Upload to your root
Place the file at https://yourdomain.com/llms.txt (or /ai.txt). Serve with Content-Type: text/plain. Done.
Frequently asked questions
llms.txt is a 2024 proposal from Jeremy Howard (Answer.AI) for a small Markdown file at the root of your domain that summarizes your site and lists canonical URLs for large language models. It complements robots.txt and sitemap.xml — while robots.txt controls who can crawl and sitemap.xml lists every URL, llms.txt curates the few pages you actually want LLMs to learn from. You should publish one if you want to be cited cleanly in ChatGPT, Claude, Perplexity or Gemini answers.
Upload it to the root of your domain so it's accessible at https://yourdomain.com/llms.txt. Use Content-Type: text/plain; charset=utf-8. For ai.txt the convention is the same path: /ai.txt. Subdomains can each have their own file. The expanded llms-full.txt lives next to it at /llms-full.txt for crawlers that prefer one big payload.
llms.txt is the index — a short Markdown file listing your most important pages with one-line descriptions. llms-full.txt is the concatenated bundle — the same structure but with the full Markdown body of each page inlined so an LLM can ingest everything in one fetch. Sites like Anthropic, Cloudflare and Stripe ship both: the small file for navigation, the big file for offline training and RAG.
As of 2026 Google has not officially adopted llms.txt for Search or AI Overviews. However ChatGPT, Claude, Perplexity, Hugging Face Smol and several open-source RAG tools do consume it. Google honors Google-Extended in robots.txt as a separate opt-out for Bard/Gemini training. Best practice today: publish llms.txt for the AI ecosystem that already reads it, plus a proper robots.txt with Google-Extended directives.
About llms.txt & Generative Engine Optimization (GEO)
In September 2024 Jeremy Howard of Answer.AI proposed llms.txt as a simple Markdown file living at the root of any web property. The premise: large language models have small context windows compared to the size of a typical website, and asking them to crawl JavaScript-heavy pages is expensive and lossy. A 2-KB curated index of your best pages, with one-sentence descriptions, fixes both problems.
Why GEO replaces (some of) SEO
Generative Engine Optimization is the practice of writing for answer engines — ChatGPT, Claude, Perplexity, Google AI Overviews — rather than the classic blue-link results page. Traditional SEO optimizes for ranking; GEO optimizes for citation. When ChatGPT writes a paragraph and links three sources at the bottom, GEO is what gets you into those three sources. The mechanism: clear writing, structured data, machine-readable summaries, and a discoverable llms.txt.
The growing ecosystem
By early 2026 the public adopters include Anthropic, Cloudflare, Stripe, Vercel, Cursor, Mintlify, Modal, and most major dev-tools companies. The llmstxt.org spec has been stable for over a year. Tools like Mintlify auto-generate llms.txt from docs; hcodx.com publishes one for its 280+ developer tools.
ai.txt vs llms.txt
They solve different problems. llms.txt is opt-in promotion — here are the pages I want you to read. ai.txt (from Spawning.ai) is opt-out — here are the AI agents I do not want training on my content. Most sites should publish both: an llms.txt that promotes the pages they care about, plus an ai.txt that opts out from training agents they don't want.