Everything you would do by hand.

Getting a website ready for AI is a list of chores. Here is the list, and beside each chore the desk that does it: on a real domain, with the price next to it.

  1. Check

    Free
    • $curl -sI https://you.com/
    • $curl https://you.com/robots.txt
    • read it for GPTBot, ClaudeBot, Google-Extended
    • $curl https://you.com/sitemap.xml
    • $curl https://you.com/llms.txt
    • view source: find the JSON-LD, the canonical, the Open Graph tags
    • read 41 specs, keep notes, do it again for every site

    AI Readiness Audit

    41 checks in 5 categories: can AI reach your site, find what matters, understand it, read it as data, and quote it. One score, every finding, the fix beside each. Free, no account.

    llmstxt.studio audited 3 Oct 2026, 20:13 UTC

    96/100

    Excellent

    • 30 passed
    • 4 warnings
    • 0 errors

    We ran 34 checks in 3.9 seconds.

    Where the score comes from

    AI Access 100

    No problems

    AI Discovery 97

    1 warning

    Content Understanding 93

    1 warning

    Machine-Readable Data 93

    1 warning

    Answer Readiness 88

    1 warning

    All 41 checks, by name Run it on a domain

  2. Read every page

    Pro
    • $curl https://you.com/sitemap.xml | grep '<loc>'
    • open each URL: title, H1, heading order, author, date, schema
    • 18 of the 41 checks are per page. Multiply by your page count.

    Deep Site Crawl

    Reads up to 100 pages a run on Pro and 500 on Premium, runs the 18 page-level checks on each, and rolls them up so you see which pages fail which check, not just that something does. 10 crawls a month on Pro, 30 on Premium.

    The 18 checks that run on every page

  3. Fix

    Free account
    • open the llms.txt spec; write the file from scratch
    • choose the key pages; write a description for each
    • fix the robots file, the sitemap, the schema, one check at a time
    • re-run everything after each change

    The fix, per finding

    Every failing check names the fix and where it goes. When the fix is an llms.txt, the llms.txt Generator writes it from your sitemap and AI Enhancement turns URL-pattern descriptions into real ones. Re-check the moment you ship. A free account includes 3 AI generations; Pro has no limit.

    What we would fix first on llmstxt.studio

    1. Warning No llms-full.txt file Your site has no llms-full.txt file.
    2. Warning Skipped heading levels Your homepage skips heading levels: h2 → h6.

    What llms.txt is Whether it works

  4. Watch

    Pro
    • every Monday:
    • $curl -s https://you.com/sitemap.xml | diff - last-week.xml
    • after every deploy: is llms.txt still there? re-run the 41 checks
    • keep a spreadsheet of scores; notice when one drops

    Audit Monitoring

    Sitemap Monitoring and Deploy Status run daily on Pro and email only when something changed. Scheduled re-audits keep 90 days of score history on Pro and 365 on Premium, and each run names the checks that flipped. On Free, monitoring is manual.

    What each scheduled run records. Your history starts with the first one.

    Score
    The number, and how far it moved since the last run
    Checks that flipped
    Each one named, pass to fail and back
    Sitemap
    Pages added or removed since the last look
    llms.txt
    Still deployed, or gone

  5. Measure

    Free · Pro
    • write down the questions your customers ask an AI
    • ask an AI search engine each one; note who it names
    • do it again next month; compare by hand

    AI Citation Check

    We write 8 queries from your llms.txt in three tiers, run them against an AI search engine, and tell you where you stand and who was named instead: one competitor a query on Free, 3 on Pro, all of them on Premium. 50 checks free; 100 a month on Pro, 500 on Premium.

    How a result reads

    1. Not Yet Visible
    2. Emerging
    3. Growing
    4. Strong
    Brand discovery
    What a customer asks when looking for what you offer.
    Topic authority
    What someone asks to learn what your site could teach them.
    Competitive landscape
    The broader questions where you compete with bigger names.

  6. Compare

    Public
    • $for d in $(cat top-20000.txt); do curl -s https://$d/llms.txt; done
    • keep every version
    • do it again next week, for a year

    llms.txt Registry

    3.99% of 999,839 ranked domains publish an llms.txt. We have been watching for 42 days, and 3,320 files changed in the last thirty.

    Every audited domain joins the registry, so your score is read against what the web actually does, not against a checklist alone. First seen, last changed, every version: nobody can backfill that record later.

    llmstxt.studio its row in the llms.txt Registry

    llms.txt
    present
    First seen
    22 Aug 2026
    File last changed
    Not seen changing
    Changes recorded
    1
    Last checked
    3 Oct 2026

    Adoption statistics Browse the registry Examples directory

  7. In the lab

    Open
    • ask an AI ten questions about your site; count the tokens it burned getting each answer

    AI Read Cost

    How much work an AI system needs to answer questions about your site, with the answer's quality shown beside the cost. Early, and open to try on any domain.

    Try AI Read Cost Read the research

Now run it on yours.

Free score, no sign-up. Results in seconds. Or compare the plans