What is llms.txt?
llms.txt is a markdown file at /llms.txt that describes your service in natural language for LLMs. While OpenAPI gives machines a structured spec, llms.txt gives language models the context they need to understand what your service does, when to use it, and how.
Format
Per the llmstxt.org spec, the file must start with an H1 heading:
# Your Service Name
> Brief one-line description of what this service does.
## Overview
Your Service lets agents search and analyze public datasets. All read
endpoints are free. Write endpoints use x402 micropayments.
## Endpoints
- GET /api/search?q=term — Search datasets (free)
- POST /api/analyze — Run analysis on a dataset ($0.10 per call)
- GET /api/status — Health check (free)
## Authentication
Free endpoints need no auth. Paid endpoints return HTTP 402 with an
x402 payment challenge. Pay with USDC on Base.
## Examples
Search for climate data:
GET /api/search?q=climate+data
Analyze a dataset:
POST /api/analyze {"dataset_id": "climate-2024", "type": "summary"}
Why it matters
LLMs process natural language better than JSON schemas. llms.txt gives them the "human context" — what the service is for, practical examples, and the big picture. It complements OpenAPI rather than replacing it.
Spec maturity
Community standard. Defined at llmstxt.org. Widely adopted — major sites including Anthropic, Cloudflare, and Stripe publish llms.txt.
Learn more
- llmstxt.org — Specification