The AI Readiness Index Open standard by TEN7

Open standard · Version 12.0

Can machines find, read, trust and use your website?

The AI Readiness Index is a public standard for answering that question: 72 tests, one hundred points, and every claim tied to someone who is not us. Free to read, to implement and to challenge.

The score

One hundred points, four questions, asked in order

Each category answers one thing a machine needs before it can use your site. The weights are published, every one is a whole number, and a change to any of them is a new major version.

  1. Reach4 tests

    Can they get to it?

    See the tests
  2. Read19 tests

    Can they find your pages and read the words?

    See the tests
  3. Understand11 tests

    Do they know what it is?

    See the tests
  4. Act4 tests

    Can an agent do something?

    See the tests

Evidence, not opinion

A test earns points only if something is documented using it

Every test carries an evidence label and graded sources, and is re-read every six months. Where we found no documented reader, the test is still measured and shown, but it is worth nothing and a site is never marked down for lacking it.

  • AI product68 pts
  • Search index25 pts
  • Practice7 pts

See every source and review date

Beside the score

Three groups that are measured and never scored

Posture

Choices that are yours to make. A choice is not a defect, so these are shown and never scored.

6 items

Extras

Good practice with no documented reader, and a few early signals. Measured and reported, never scored.

20 items

Measurement

How you track AI traffic. Real work that a scan cannot see, so it is guidance, not points.

8 items

Take part

The standard is changed in public

Propose a test, correct a citation or challenge a weight. Every accepted change is recorded in the changelog, and anything that moves a score is a new major version that says plainly old and new numbers cannot be compared.

Propose a test

Describe what a machine needs, the signal that shows it and a source that is not us. Open an issue.

Open an issue

Talk it through

Questions, disagreements and half-formed ideas belong in Discussions, before they are proposals.

Join the discussion

Send a change

The whole standard is one Markdown file and one registry. Pull requests are checked automatically.

How contributing works

About this standard

An open standard, not a product

The AI Readiness Index was created by TEN7 and is published here under a license that lets anyone copy it, adapt it and build a scanner from it. It does not belong to any scanner. A score is a number you can reproduce from the text on this site.

TEN7 is the company behind the standard, and readinessindex.io is a product of TEN7: a scanner that implements it. They are linked so you can see where the standard came from, not to recommend them. Read more about how it is run.

  • TEN7The company that created the standard and maintains it with outside contributors.
  • readinessindex.ioA product of TEN7: a scanner that implements the standard. It is one implementation, not the standard.