llms.txt for WordPress — What It Is, How to Add It, Why It Matters

What is llms.txt? llms.txt is a plain-text file at the root of a website (e.g. yoursite.com/llms.txt) that helps AI assistants and large-language-model crawlers understand and cite the site’s content. It functions as a curated map of the site’s most important pages, written in a format optimised for LLM consumption rather than for traditional search-engine crawling. The proposal originated from Jeremy Howard (Answer.AI) in late 2024 and has been adopted by a growing number of documentation sites, technical product pages, and content-first publishers.

Does my WordPress site need llms.txt? If your content is being read by users through ChatGPT, Claude, Perplexity, or Google AI Overviews — and you want those assistants to cite your site accurately when they reference your topics — then yes, llms.txt is worth shipping. It’s not (yet) a ranking signal in the classic Google sense, but it’s a citation signal for the generative-search layer that’s increasingly mediating how users find content.

How do I add llms.txt to WordPress? Two paths: (1) Manually — create a llms.txt file with the structure shown below, upload it to your site root via SFTP or your host’s file manager. (2) Automatically — install a plugin that generates and maintains llms.txt from your site’s content. Asteris for WordPress’s SEO + AI module includes this; the file is generated, schema-aware, and refreshes on a schedule. See the SEO + AI module →

What llms.txt looks like

A minimal llms.txt file follows this structure:

# Your Site Name

> One-sentence description of what the site does.

Optional longer description with context an LLM should know before
reading the site (audience, scope, key terminology).

## Section heading

- [Page title](https://yoursite.com/url): Brief description of what this page covers.
- [Another page](https://yoursite.com/another): Description.

## Another section

- [Page](https://yoursite.com/page): Description.

The spec is deliberately simple: Markdown, a single H1 for the site name, a blockquote tagline, optional paragraph context, then H2 sections containing bulleted lists of links with descriptions.

A companion file, llms-full.txt, contains the full text of the listed pages concatenated. This lets an LLM consume the entire site’s content corpus without crawling individual URLs.

How llms.txt differs from robots.txt and sitemap.xml

These three files coexist; they serve different purposes.

• robots.txt controls access. sitemap.xml provides exhaustive coverage. llms.txt provides curated importance.
• sitemap.xml is for the bot, machine-to-machine. llms.txt is for the bot, optimised for an LLM to actually read and reason about.
• An ideal site has all three.

Does llms.txt improve visibility in ChatGPT, Claude and Perplexity?

Honestly: the evidence is still early. llms.txt adoption is recent enough (late 2024-2025) that we don’t yet have a year of citation data showing “sites with llms.txt get cited X% more often”. The mechanism is plausible — LLMs work better with curated context than with raw site-crawl noise — but the empirical proof is thin.

What we do know:

• Anthropic, Cloudflare, and other developer-platform companies have published llms.txt files publicly.
• AI assistants are increasingly trained to look for curated content surfaces.
• The cost of shipping llms.txt is near-zero (one file, auto-generated). The upside is asymmetric.

Our recommendation: ship it. The cost is trivial; the differentiated visibility is real even if the magnitude is uncertain.

Block AI bots vs allow citation crawlers — what should you do?

Some site owners block AI crawlers entirely (via robots.txt User-agent: GPTBot → Disallow: /) to prevent their content being used as training data. Others allow them to encourage citation.

The honest split: training-class crawlers (which ingest content into model training datasets) and citation-class crawlers (which read content live to ground answers in real-time) are increasingly separate user-agents. You can allow one and block the other.

Our recommendation for most content sites: allow citation-class crawlers (you gain citation visibility), block training-class crawlers (you preserve your IP from being absorbed without attribution). Asteris’s AI bot blocker lets you make this distinction per-bot in one settings panel.

How to add llms.txt to WordPress

Option 1: Manually

1. Open your text editor of choice.
2. Write a llms.txt following the structure shown above. The H1 is your site name. The blockquote is a one-sentence description. The sections list your most important pages.
3. Save as llms.txt (UTF-8, LF line endings).
4. Upload to your site’s root directory via SFTP, your host’s file manager, or your hosting control panel. The file must be accessible at https://yoursite.com/llms.txt — not /wp-content/llms.txt or any subdirectory.
5. Verify by opening https://yoursite.com/llms.txt in a browser. You should see the raw text.

Limitation: you’ll need to manually update the file every time you add or rename important content. For sites with more than ~20 pages, this becomes impractical quickly.

Option 2: Automatically (with Asteris)

1. Install Asteris for WordPress (free version or a paid tier — the SEO + AI module’s llms.txt generator is in the paid module).
2. WP Admin → Asteris → Modules → toggle SEO + AI Suite to ON.
3. Asteris → SEO + AI → Tools → llms.txt → enable.
4. Asteris generates llms.txt (and llms-full.txt) from your content automatically. It refreshes on a daily schedule by default. The file is served at yoursite.com/llms.txt via a WordPress rewrite rule (no actual file on disk).
5. Customise the auto-generated content if needed — site name, tagline, section structure, which post types to include — in the same admin panel.

Option 3: Other plugins

A handful of dedicated llms.txt plugins exist on WordPress.org (e.g. “Website LLMs.txt”, “AEOmatic”). They generate the file but don’t integrate with broader SEO + AI bot management. If you only need the file, they work fine. If you also need AI bot blocking, AI traffic tracking, and the full SEO toolkit, Asteris bundles it.

Eat your own dog food

This site (asterisforwordpress.com) ships its own llms.txt. View it at asterisforwordpress.com/llms.txt — it’s generated by the same Asteris SEO + AI module described above, so you can see the output before installing. (The companion llms-full.txt is at /llms-full.txt.)

How llms.txt fits into Generative Engine Optimization (GEO)

llms.txt is one piece of a broader practice that’s emerged in 2024-2025: Generative Engine Optimization — the discipline of making content visible and citable inside AI-driven answer engines, in the same way SEO makes content visible inside link-driven search engines.

GEO overlaps with SEO but extends into:

• Structured, citable claims (one-sentence facts AI can quote)
• Definitional openers (lead the page with a clear definition)
• Schema markup (machine-readable context)
• llms.txt (curated content surface)
• AI bot management (control what trains on you)
• Tracking AI-driven referral traffic

If you’re investing in SEO in 2026, GEO is the adjacent surface — and llms.txt is the cheapest, lowest-friction first step.

Frequently asked questions

What is llms.txt? A plain-text Markdown file at the root of a website that helps AI assistants understand and cite the site’s content. It curates the most important pages with descriptions, formatted for LLM consumption.

How do I add llms.txt to WordPress? Either manually (write the file, upload to site root) or automatically (install a plugin like Asteris for WordPress’s SEO + AI module that generates and maintains it).

Is llms.txt different from robots.txt and sitemap.xml? Yes. robots.txt controls crawler access. sitemap.xml provides exhaustive URL coverage. llms.txt curates the most important pages for LLM understanding. They serve different purposes; an ideal site has all three.

Does llms.txt improve visibility in ChatGPT, Claude and Perplexity? The evidence is still early. The mechanism is plausible and adoption is rising, but empirical citation-rate data is thin. The cost of shipping llms.txt is near-zero; the upside is asymmetric.

Should I block GPTBot and ClaudeBot from my WordPress site? Depends on goals. To preserve IP from training, block training-class crawlers (GPTBot, ClaudeBot, Google-Extended, CCBot). To gain citation visibility, allow citation-class crawlers (ChatGPT-User, OAI-SearchBot, PerplexityBot, Anthropic-AI). You can do both — Asteris’s AI bot blocker manages per-bot.

What is llms-full.txt? A companion file containing the full text of every page listed in llms.txt, concatenated. Lets an LLM consume the entire site’s curated corpus in one fetch.

See the SEO + AI module → · WordPress SEO pillar → · Asteris vs Yoast → · Pricing →