llms.txt vs agents.txt
The map lists them as complementary: sites and agents commonly implement both. They answer different questions: llms.txt is a step 4 (Read) standard, agents.txt step 1 (Discover), so they are not substitutes.
| llms.txt | agents.txt | |
|---|---|---|
| Lifecycle step | 4 Read: Give me the content | 1 Discover: What is here, and where? |
| Answers | Where can an agent find a concise, curated map of a site's content, written for language models rather than browsers? | Which agent-facing protocols (payments, auth, MCP, skills, A2A) does this site speak, and where? |
| Governed by | Jeremy Howard / Answer.AI; maintained as an open proposal in the AnswerDotAI/llms-txt GitHub repository with a public Discord for implementation discussion | None; at least three non-converging proposals (agents-txt.com, kaylacar/agents-txt, agentstxt.dev) each self-describe as "the" agents.txt standard, with no shared maintainer or reconciliation process found |
| Status (issuer's words) | "This is v2 of the proposal" — explicitly a community proposal open for input, not a standards-track document | "An open standard proposal for declaring what AI agents can do on a website" (kaylacar/agents-txt); a Mastercard developer-documentation page independently characterizes the space as "IETF draft expired ... fragmented landscape with 11+ competing proposals ... no ratified standard" |
| Version | v2 (2026-08) | agents-txt.com: v1.0; competing proposals carry no shared version |
| Form / transport | well-known file | well-known file |
| Found at | /llms.txt (or /<path>/llms.txt, e.g. /docs/llms.txt) | /agents.txt (well-known root path per all identified proposals; one proposal also allows /.well-known/agents.txt) |
| Implementation | multi-implementation — Rankability (Tranco top-1000, June 2026: 8.7%) and SE Ranking (~300,000 domains analyzed: 10.13%) — independent third-party publisher-side counts, not our own probe | prototype |
| Evidence | 4 verified, 4 reported | 3 verified, 1 reported |
| Primary use case | Serve readable content to agents | Let agents discover my API or service |
| Last verified | 2026-10-23 (current) | 2026-10-23 (current) |
| Specification | https://llmstxt.org | https://agents-txt.com/spec |
When to use which
- Choose llms.txt when your question is “Where can an agent find a concise, curated map of a site's content, written for language models rather than browsers?” and you accept its maturity: status ""This is v2 of the proposal" — explicitly a community proposal open for input, not a standards-track document", implementation multi-implementation (Rankability (Tranco top-1000, June 2026: 8.7%) and SE Ranking (~300,000 domains analyzed: 10.13%) — independent third-party publisher-side counts, not our own probe).
- Choose agents.txt when your question is “Which agent-facing protocols (payments, auth, MCP, skills, A2A) does this site speak, and where?” and you accept its maturity: status ""An open standard proposal for declaring what AI agents can do on a website" (kaylacar/agents-txt); a Mastercard developer-documentation page independently characterizes the space as "IETF draft expired ... fragmented landscape with 11+ competing proposals ... no ratified standard"", implementation prototype.
Read the full articles
- llms.txt: Where can an agent find a concise, curated map of a site's content, written for language models rather than browsers?
- agents.txt: Which agent-facing protocols (payments, auth, MCP, skills, A2A) does this site speak, and where?