Agent Native

Copy page

llms.txt

llms.txt is a proposed standard for making website content accessible to large language models. Similar to robots.txt for web crawlers, llms.txt tells AI tools what documentation is available and how to access it.

Files

Binance API Documentation provides two llms.txt files:

File	URL	Description
llms.txt	/dev-docs/llms.txt	Summary — document titles, descriptions, and links
llms-full.txt	/dev-docs/llms-full.txt	Full content — complete documentation in a single file

Format

llms.txt (Summary)

The summary file contains a structured list of all documentation pages with titles, descriptions, and URLs:

Code
# Binance API Documentation

> Comprehensive API documentation for Binance trading platform.

## Spot Trading

- [Changelog](/docs/spot-trading/1.changelog): Spot Trading API changelog
- [README](/docs/spot-trading/2.readme): Getting started with Spot Trading API
...

## Derivatives Trading

- [Change Log](/docs/derivatives/change-log): Derivatives API changelog
...

llms-full.txt (Full Content)

The full file includes the complete markdown content of every documentation page, separated by headers. This is useful for loading the entire documentation into an LLM's context window.

Usage Examples

Feed into any LLM

Code
# Download and use as context
curl -s https://developers.binance.com/en/dev-docs/llms.txt

# Full documentation for comprehensive context
curl -s https://developers.binance.com/en/dev-docs/llms-full.txt

Use with Claude

Code
Please read https://developers.binance.com/en/dev-docs/llms.txt and help me
understand how to place a futures order using the Binance API.

Use in a Python script

Code
import requests

# Fetch the documentation index
response = requests.get("https://developers.binance.com/en/dev-docs/llms.txt")
docs_index = response.text

# Feed into your LLM as context
messages = [
    {"role": "system", "content": f"Binance API docs:\n{docs_index}"},
    {"role": "user", "content": "How do I check my account balance?"}
]

Use with LangChain

Code
from langchain_community.document_loaders import WebBaseLoader

loader = WebBaseLoader("https://developers.binance.com/en/dev-docs/llms-full.txt")
docs = loader.load()

When to Use llms.txt vs Other Options

• llms.txt — You want to quickly give an LLM awareness of all available documentation, or load full docs into context
• Agent REST API — You need structured JSON data, filtered queries, or programmatic access
• MCP Server — You're using Claude Code, Cursor, or another MCP-compatible IDE