llms.txt for Therapists: The 10-Minute AI Visibility Signal — An Honest 2026 Guide
A plain Markdown file placed at the root of your website can help AI systems understand your therapy practice faster and more accurately. Here's what llms.txt actually is, whether it measurably increases AI citations, and how to write one for your practice in about 10 minutes.
What Is llms.txt? (And Why Therapists Are Hearing About It Now)
Nearly half of consumers — 45% — now use AI tools when looking for local business recommendations, according to BrightLocal's 2026 Local Consumer Review Survey. A year ago the figure was 6%. Among adults aged 30 to 44, it has already crossed 64%. That pace of adoption is why a 2024 web standard called llms.txt has started appearing in every marketing newsletter you're reading — and why it's worth understanding before you decide whether your therapy practice needs one.
llms.txt is a plain text file, written in Markdown, placed at the root of your website so it's accessible at yourdomain.com/llms.txt. The format was proposed in September 2024 by Jeremy Howard, co-founder of Answer.AI and fast.ai, to address a concrete problem: AI language models have limited context windows and struggle to parse the dense HTML and JavaScript of modern websites efficiently (Howard, Answer.AI, 2024). A typical therapy website layers navigation menus, cookie banners, sidebars, and footer boilerplate on top of the actual content an AI needs. llms.txt cuts through that and hands the model a clean, structured summary directly.
Think of it as a curated table of contents written for AI crawlers rather than browsers: your practice name, what you do, where you practice, and links to the pages that matter most — in a format a language model can consume without inference or guesswork.
How Is llms.txt Different from robots.txt?
If you've read our guide to AI crawlers and robots.txt, you know robots.txt controls which bots can access your site and what they're permitted to do — index for search, train on your content, or neither. llms.txt is entirely different in purpose.
robots.txt answers: "Which parts of my site can this crawler visit?"
llms.txt answers: "Once you're in, here's what my site is about and which pages matter most."
The two files are complementary, not competing. robots.txt keeps training bots off your clinical content if you choose. llms.txt helps AI search bots understand your practice quickly, accurately, and without hallucinating. You can — and in most cases should — have both.
There is also a companion file called llms-full.txt. Where llms.txt links to your key pages for an AI to follow, llms-full.txt contains the full Markdown text of those pages inline — eliminating the need for the AI to make additional requests. For most therapy practices, the basic llms.txt is sufficient to start.
What Does llms.txt Do for a Therapy Practice?
When an AI crawler is programmed to read your llms.txt file, it gains three practical advantages in understanding your practice:
- An accurate first impression. Instead of inferring what you do from fragments of HTML — a bio paragraph here, a services list there — the AI gets a clean, structured summary of who you are, what you treat, and where you practice. A therapist whose specialty is EMDR for complex trauma is far less likely to be misdescribed if she has stated that explicitly in a machine-readable file.
- Priority signals. You choose which pages appear in your llms.txt. Your top specialty page, your About page, your FAQ — you tell the AI these are the pages worth reading, rather than leaving it to guess by crawl order or link weight.
- Reduced hallucination risk. AI systems are more likely to represent your practice accurately when they have unambiguous, authoritative source data. This matters practically: as conversational AI becomes the first contact for clients seeking mental health support, a misrepresented specialty or wrong city can mean a misdirected referral.
Does llms.txt Actually Increase AI Citations? The Honest Answer
The straightforward answer is: probably not yet, but the cost of implementing it is genuinely close to zero.
As of June 2026, about 8.7% of the world's top 1,000 websites publish an llms.txt file (Rankability, June 2026). Adoption is growing — Yoast SEO began auto-generating the file for WordPress sites, signaling mainstream uptake — but the evidence on citation impact is thin. No major AI company has publicly committed to reading llms.txt as an official standard, and correlation analyses across large samples of domains have not found a measurable citation-lift from having the file. Anthropic and Perplexity have signaled they respect the format; others have not confirmed this publicly.
The counterpoint is the implementation cost. Writing an llms.txt for your therapy practice takes roughly 10 minutes. The file is a handful of lines of Markdown. If AI crawlers begin reading it at scale as the standard matures — which is plausible as Anthropic, Perplexity, and others have expressed intent — you've already done the work. If they don't read it this year, you've invested 10 minutes.
For therapists, there's also an accuracy benefit that matters independent of citation counts: providing clean, structured information about your practice reduces the likelihood that AI systems misrepresent you to a prospective client. That's a real concern, and it's one llms.txt directly addresses.
What matters more right now: The signals covered in The Therapist's Guide to AI Discoverability — structured schema markup, Google Business Profile, consistent NAP across directories, and content quality — have measurable, documented impact on AI citations today. Do those first. Then, because it takes 10 minutes, do llms.txt too.
How to Write an llms.txt for Your Therapy Practice
Create a plain text file named exactly llms.txt (not llm.txt, not LLMs.txt — the exact casing matters for some parsers) and place it at your domain root so it's accessible at yourdomain.com/llms.txt. The file should return a 200 status with plain text or Markdown content type.
The required elements per the Jeremy Howard proposal are:
- An H1 with your project or practice name
- A blockquote with a short, one-paragraph summary of your practice containing the key information an AI needs
- One or more H2 sections containing links to key pages with brief one-line descriptions
- An optional "Optional" H2 section for lower-priority pages the AI can skip if context is tight
Template for a solo therapy practice:
# [Practice Name]
> [Therapist Name], [Credential], is a licensed therapist in [City, State]
> specializing in [top 2-3 specialties]. [Practice Name] offers
> [in-person/telehealth/both] therapy for [client population, e.g., adults,
> couples, adolescents]. Phone: [number] | [email]
## Key Pages
- [About](https://yoursite.com/about): Therapist bio, credentials, training, and clinical approach
- [Services](https://yoursite.com/services/emdr): EMDR therapy for trauma and PTSD
- [Services](https://yoursite.com/services/anxiety): Anxiety and panic treatment
- [FAQ](https://yoursite.com/faq): Insurance, fees, and what to expect
- [Contact](https://yoursite.com/contact): Office location, telehealth access, scheduling
## Optional
- [Blog](https://yoursite.com/blog): Articles on anxiety, depression, trauma, and relationship health
Tips specific to therapy practices:
- Include your credential in the blockquote — LPC, LCSW, LMFT, PhD, PsyD. AI systems weight credentials as authority signals for healthcare content, and your license type is one of the first things a prospective client asks about.
- Name your specialties explicitly. "Specializing in anxiety, EMDR for complex trauma, and couples therapy" is more useful to a language model than "a wide range of mental health concerns."
- Include your city and state. Most clients find therapists through local searches, and geographic context is essential for AI local recommendations to be accurate.
- Link to individual specialty pages rather than a general services page, if you have them. A dedicated EMDR page tells an AI more than a single page that lists all modalities together.
- Keep it current. If you move offices, change your phone number, or stop accepting a specific insurance, update llms.txt as part of that same change. Stale information in a file designed to reduce hallucinations creates the exact problem it's meant to prevent.
Where llms.txt Fits in Your AI Visibility Stack
llms.txt is a low-cost complement to the signals that actually drive AI citation rates for therapists today. Here's how it fits relative to other discoverability actions:
| Signal | Impact on AI citations | Estimated effort |
|---|---|---|
| Google Business Profile — claimed, accurate, current | High | 30–60 min to claim; ongoing updates |
| Foursquare listing — claimed, NAP matching your website | High | 15 min |
| Structured data — MedicalBusiness, Person, FAQPage schema | High | Requires technical implementation |
| AI crawlers explicitly allowed in robots.txt | High | 5 min |
| Google Reviews — 10 or more, with recent activity | High | Ongoing |
| Content with question-format headings and cited sources | Medium-high | Per post or page |
| llms.txt | Emerging — no proven citation lift yet | 10 min |
The order of that table reflects the order to prioritize your time. The high-impact signals are established, well-documented, and actively influence what ChatGPT, Perplexity, Google AI Overviews, and Apple Intelligence return today. llms.txt is worth doing because of its negligible cost and because the standard is actively being adopted — not because it will move the needle on its own.
If you've already worked through the high-impact signals and want to make sure every available layer of AI context is in place, llms.txt is the natural next step. If you're still building out your Google Business Profile or haven't claimed your Foursquare listing, do those first.
Ready to handle the foundational signals without managing a technical checklist? WebsiteTherapy builds structured data, AI crawler access, Foursquare monitoring, IndexNow indexing, and ongoing AI visibility tracking into every client site from launch. See plans and pricing.
Sources: BrightLocal Local Consumer Review Survey 2026 (brightlocal.com/research/lcrs-ai-trust/); Jeremy Howard / Answer.AI, "llms.txt proposal," September 2024 (answer.ai/posts/2024-09-03-llmstxt.html); Rankability, "LLMS.txt Adoption: 8.7% of the Top 1,000," June 2026 (rankability.com/data/llms-txt-adoption/).