Ecosystem evidence

verified-as-of: 2026-08-30 (every number below is dated; this domain moves monthly)

Contents

  1. The one-line verdict
  2. Adoption measurements, dated
  3. Who publishes
  4. Who reads — server-log studies
  5. Google and the “is it dead” thread
  6. Directories and registries
  7. Vendor sources graded
  8. References

1. The one-line verdict

Adoption is real and growing (≈5–10% of the general web by mid-2026, 28% among SEO-savvy sites, 8.8× year on year); unsolicited consumption is near zero (97% of files never get an AI request); the demonstrated use is agents that are pointed at the file — the Claude-Code UA out-fetched every AI retrieval bot bar two (statespace-indexer, GPTBot).12 Publish one for agents and coding tools; do not expect citations or rankings from it.3

2. Adoption measurements, dated

DateSourceSampleFinding
Feb→May 2025Chris GreenMajestic Million15 → 105 valid files (~0.01%); ~100k crawl errors caveat4
Jun 2025Originality.ai3M+ sites4,088 llms.txt5
Jun 2025RankabilityTranco top 1,0000.3%6
Jul 2025HTTP Archive (Burridge)top 10k1.04% valid7
Nov 2025SE Ranking~300k domains10.13% overall (9.88% low-traffic / 10.54% mid / 8.27% 100k+ visits)3
Mar 2026Originality.ai via ppc.landFortune 5007.4% (37/500)8
May 2026Originality.ai3M+ sites36,120 llms.txt (8.8× YoY); llms-full.txt 23 → 2,463 (107×); ai.txt 3975
May 2026Ahrefs137,210 Ahrefs-Web-Analytics domains28% publish a valid file (self-selected sample)1
Jun 2026HTTP Archive (Burridge)top 1k / 10k / 100k / 1M6.28% / 5.61% / 5.17% / 5.07% (~5.4× in 12 months)7
Jun 2026RankabilityTranco top 1,0008.7% (87 files; 15 with llms-full.txt)6

Contradictions kept: top-1000 adoption reads 6.28% (HTTP Archive/Chrome list) vs 8.7% (Tranco) for the same month;67 third parties attribute “0% in the top 1000” to SE Ranking, whose primary article gives no such figure;9 Ahrefs’ 28% is not comparable with population figures because of sample bias;1 a “51.8% of a 219-host panel” claim (Presenc.ai, Aug 2026) has no supporting data on its page.10

3. Who publishes

4. Who reads — server-log studies

StudyWindow / sampleFinding
Ahrefs (2026-06-15)May 2026 logs, 137,210 domains97% of valid files got zero requests; of requests, 96% bots, 77% of those non-AI (SEO auditors 21.7%); named AI bots 19.5%; AI training crawlers 5.3% (GPTBot 4.51%, ClaudeBot 0.8%); AI retrieval 1.1% (OAI-SearchBot 0.74%); 0 AI requests to non-existent files (nobody probes speculatively); the Claude-Code UA out-fetched every AI retrieval bot bar statespace-indexer and GPTBot1
OtterlyAI (2026-02-05)90 days, one site84 of 62,100 AI-bot requests hit /llms.txt (0.1%)15
Wislr (Feb–Mar 2026)48 days, one site12,099 bot requests; robots.txt fetched hundreds of times (OAI-SearchBot 180, ClaudeBot 175); sitemap.xml too; llms.txt 016
EZY Research (Apr–Jul 2026)83 sites, 12 weeksrobots vs llms: GPTBot 3,990/7, ClaudeBot 3,120/9, PerplexityBot 775/0, Googlebot 5,125/67, Meta-ExternalAgent 172/193 (the only bot fetching it more than robots.txt)17
Hacker News thread (Feb 2026)anecdotal logsonly OVH/GCP-hosted tools (WebPageTest, BuiltWith), no ChatGPT/Claude UAs18
Cloudflare Accept: text/markdown (Mar–Apr 2026)44 days, one Worker1,421 requests: headless Chrome 639, “Claude” (Anthropic infra) 500, axios 211; no GPTBot/PerplexityBot/ClaudeBot19

Citation impact: SE Ranking’s 300k-domain model (Spearman + XGBoost + SHAP, 2025-11-07) found no relationship between having an llms.txt and LLM citation frequency — removing the variable improved model accuracy.3 Correlational; which LLMs’ citations were measured is unspecified.

5. Google and the “is it dead” thread

6. Directories and registries

Self-submitted, overlapping, unverified — lower bounds, not measurements:27282930

DirectorySizeNotes
directory.llmstxt.cloud“4,000 websites listed” (49M llms.txt tokens / 325M llms-full tokens)named in spec v2
llmstxthub.com~2,650 entries, 15–16 categories (David Dias)named in spec v2
llmstxt.site~1,000+ (≈170 in May 2025); columns product / website / llms.txt / llms-full.txt / token counts; /submitnamed in spec v2
SecretiveShell/Awesome-llms-txt784 link lines (counted 2026-08-30)GitHub
llms-text.com“780+ verified implementations”vendor’s own directory

None publishes an llms.txt-of-llms.txt; llmstxt.site’s token-count column is the most useful signal for consumers budgeting context.29

7. Vendor sources graded

PageAuthor / dateClaimsGrade
llms-text.com/blog/sites-using-llms-txtMichael Vereb, 2025-07-25“780+ verified”; names Anthropic, Cloudflare, Supabase, Vercel, ElevenLabs, Firecrawl, Mintlify, Cursor, Aptos, GitBook, Wix; “no e-commerce adoption”adopters check out on live probe; count uncorroborated — low for numbers, fine for examples31
llms-text.com/blog/what-is-llms-txtsame“foundational pillar of GEO”; ChatGPT/Perplexity/Cursor/Windsurf/Claude Code consume it; “up to 114% more tokens” (incoherent arithmetic), “10–15% accuracy” — unattributedGEO and ChatGPT/Perplexity-consumption claims contradicted by every log study — low32
llms-text.com/blog/llms-txt, /how-to-create-llms-txtsameMIME/200/UTF-8 rules; Link: …; rel="describedby"; “under 10 KB”; framework snippets; funnels to its generator/validatoruseful mechanics (the describedby relation is now in spec v2), vendor numbers — medium3334
gitdoc.ai/blog/llms-txt-ai-readable-documentationYadian Llada / GitDoc, 2026-05-22GitBook: 41% of docs page requests from AI agents (unverified); permission / inventory / navigation distinction; curate 10–20 pages (quickstart, auth, per-resource reference, errors, changelog); regenerate in the build; llms-full for priority pagessound guidance, unverified headline stat, product promotion — medium35

References

Footnotes

  1. https://ahrefs.com/blog/llmstxt-study/ — 137,210-domain log study, 2026-06-15 (study) ↩ ↩2 ↩3 ↩4

  2. https://caseyrb.com/blog/state-of-llms-txt-adoption/ — HTTP Archive, 2026-06-20 (study) ↩

  3. https://seranking.com/blog/llms-txt/ — 300k-domain adoption + citation model, 2025-11-07 (study) ↩ ↩2 ↩3

  4. https://www.chris-green.net/post/million-websites-in-search-of-llms-txt (study) ↩

  5. https://originality.ai/blog/llms-txt-tracking-study (study) ↩ ↩2 ↩3

  6. https://www.rankability.com/data/llms-txt-adoption/ (study) ↩ ↩2 ↩3 ↩4

  7. https://caseyrb.com/blog/state-of-llms-txt-adoption/ (study) ↩ ↩2 ↩3 ↩4

  8. https://ppc.land/llms-txt-adoption-rises-8-8x-but-97-of-files-get-zero-ai-requests/ (blog) ↩

  9. https://inite.ai/en/blog/is-llms-txt-dead-2026 (blog) ↩ ↩2 ↩3

  10. https://presenc.ai/research/state-of-llms-txt-2026 (vendor) ↩

  11. https://llmstxt.org/ — v2, modified 2026-08-10 (spec); live probes 2026-08-30 ↩

  12. https://www.mintlify.com/blog/the-value-of-llms-txt-hype-or-real (vendor) ↩ ↩2

  13. https://www.mintlify.com/blog/what-is-llms-txt (vendor) ↩ ↩2

  14. https://www.shopifreaks.com/shopify-quietly-rolls-out-native-llms-txt-files-for-stores-adding-structured-data-layer-for-ai-agents/ (blog) ↩

  15. https://otterly.ai/blog/the-llms-txt-experiment/ (study) ↩

  16. https://www.wislr.com/articles/ai-bot-behavior-log-analysis/ (study) ↩

  17. https://www.ezy.ai/research/do-ai-bots-read-llms-txt (study) ↩

  18. https://news.ycombinator.com/item?id=47058870 (forum) ↩

  19. https://suganthan.com/blog/cloudflare-markdown-for-agents/ (blog) ↩

  20. https://www.searchenginejournal.com/google-says-llms-txt-comparable-to-keywords-meta-tag/544804/ (docs) ↩

  21. https://searchengineland.com/google-says-normal-seo-works-for-ranking-in-ai-overviews-and-llms-txt-wont-be-used-459422 (docs) ↩

  22. https://developers.google.com/search/docs/appearance/ai-features (docs) ↩

  23. https://www.365i.co.uk/news/2025/12/09/google-llms-discover-ai-mode-2025/ (blog) ↩

  24. https://searchengineland.com/google-llms-txt-chrome-lighthouse-478246 and https://developer.chrome.com/docs/lighthouse/agentic-browsing/llms-txt (docs) ↩

  25. https://searchengineland.com/no-llms-txt-is-not-the-new-meta-keywords-458199 (blog) ↩

  26. https://ahrefs.com/blog/what-is-llms-txt/ (blog) ↩

  27. https://directory.llmstxt.cloud/ (vendor) ↩

  28. https://llmstxthub.com/ (vendor) ↩

  29. https://llmstxt.site/ (vendor) ↩ ↩2

  30. https://github.com/SecretiveShell/Awesome-llms-txt (forum) ↩

  31. https://www.llms-text.com/blog/sites-using-llms-txt (vendor) ↩

  32. https://www.llms-text.com/blog/what-is-llms-txt (vendor) ↩

  33. https://www.llms-text.com/blog/llms-txt (vendor) ↩

  34. https://www.llms-text.com/blog/how-to-create-llms-txt (vendor) ↩

  35. https://gitdoc.ai/blog/llms-txt-ai-readable-documentation (vendor) ↩