# The AI-Readability Index

We fetched the public files of 45 well-known US companies and checked three things: whether they publish an llms.txt, whether any plain-text version of their homepage exists, and whether their robots.txt lets GPTBot, ClaudeBot and PerplexityBot in. Measured 02 August 2026.

## What we found

- **12 of 45** publish an llms.txt
- **1 of 45** serve a plain-text or markdown version of their homepage
- **1** block at least one of the three AI crawlers in robots.txt

These are companies with billion-dollar revenues. The text an assistant would need to describe them mostly does not exist in a form it can take.

## Full results

| Company | Sector | llms.txt | Plain-text page | Allows GPTBot | Allows ClaudeBot | Allows PerplexityBot |
|---|---|---|---|---|---|---|
| Asana | Work management | yes | yes | yes | yes | yes |
| Amplitude | Analytics | yes | no | yes | yes | yes |
| Braze | Marketing | yes | no | yes | yes | yes |
| Datadog | Observability | yes | no | yes | yes | yes |
| Fastly | Edge network | yes | no | yes | yes | yes |
| Klaviyo | Marketing | yes | no | yes | yes | yes |
| MoneyLion | Consumer finance | yes | no | yes | yes | yes |
| Payoneer | Payments | yes | no | yes | yes | yes |
| Remitly | Remittances | yes | no | yes | yes | yes |
| Smartsheet | Work management | yes | no | yes | yes | yes |
| Trex | Decking | yes | no | yes | yes | yes |
| Twilio | Communications API | yes | no | yes | yes | yes |
| Cava | Restaurants | no | no | yes | yes | yes |
| Chewy | Pet retail | no | no | yes | yes | yes |
| Cintas | Uniforms & services | no | no | yes | yes | yes |
| Confluent | Data streaming | no | no | yes | yes | yes |
| Crocs | Footwear | no | no | yes | yes | yes |
| DocuSign | E-signature | no | no | yes | yes | yes |
| Duolingo | Education | no | no | yes | yes | yes |
| Etsy | Marketplace | no | no | yes | yes | yes |
| Fastenal | Industrial supply | no | no | yes | yes | yes |
| Grainger | Industrial supply | no | no | yes | yes | yes |
| HashiCorp | Infrastructure | no | no | yes | yes | yes |
| Hims & Hers | Telehealth | no | no | yes | yes | yes |
| La-Z-Boy | Furniture | no | no | yes | yes | yes |
| Marqeta | Card issuing | no | no | yes | yes | yes |
| Peloton | Fitness | no | no | yes | yes | yes |
| Planet Fitness | Fitness | no | no | yes | yes | yes |
| Pool Corp | Pool supply | no | no | yes | yes | yes |
| Roku | Streaming | no | no | yes | yes | yes |
| Rollins | Pest control | no | no | yes | yes | yes |
| Sleep Number | Mattresses | no | no | yes | yes | yes |
| SoFi | Consumer finance | no | no | yes | yes | yes |
| Solo Brands | Outdoor | no | no | yes | yes | yes |
| Sonos | Audio | no | no | yes | yes | yes |
| Sweetgreen | Restaurants | no | no | yes | yes | yes |
| Texas Roadhouse | Restaurants | no | no | yes | yes | yes |
| Toro | Turf equipment | no | no | yes | yes | yes |
| Traeger | Grills | no | no | yes | yes | yes |
| Watsco | HVAC distribution | no | no | yes | yes | yes |
| Wayfair | Home retail | no | no | yes | yes | yes |
| Weber | Grills | no | no | yes | yes | yes |
| Wingstop | Restaurants | no | no | yes | yes | yes |
| YETI | Outdoor gear | no | no | yes | yes | yes |
| Garmin | Devices | no | no | yes | yes | no |

## Method

For each domain we request /llms.txt, /robots.txt and the homepage. An llms.txt only counts when it returns plain text — several sites redirect that path to their HTML homepage, which is not the same thing. A plain-text page counts when the homepage advertises a markdown alternate, or /index.md returns non-HTML content. robots.txt is parsed for a root-level Disallow against each named crawler.

Reproduce it yourself: the checks are three HTTP requests per domain. Source data: https://pixelsforbots.com/ai-readiness

---

Published by https://pixelsforbots.com — a directory that gives any company a machine-readable page for $1 a cell, with a public count of AI-crawler fetches.