# Readmd

> Web page to clean, LLM-ready Markdown for AI agents: main-content extraction, metadata, outline, token counts and RAG chunks. $0.001 per page, $0.005 for 10. Readmd turns a URL into Markdown an LLM can use directly: the article without the page chrome, with metadata, an outline, a token estimate and optional RAG chunks. Plain HTTP fetch (no headless browser), robots.txt respected with the user-agent token `readmd`, a 3 MB page cap, redirects re-checked, private addresses refused. You pay only for pages that were actually read: every failure is returned unsettled. Content belongs to its publishers; Readmd fetches it on your request and does not store it. Pages that need JavaScript to render are not supported.

Payment: x402 v2, USDC on eip155:8453, scheme "exact", payTo 0x741f78ED943C671e25405dAAaAB7b6886C7e8feb. Call a paid endpoint without payment to receive HTTP 402 with the requirements (PAYMENT-REQUIRED header and JSON body); an x402 client (for example @x402/fetch) pays and retries automatically. You are charged only for successful (2xx) responses: invalid input, oversize input and upstream failures are free.

## Paid endpoints
- GET https://readmd.canisell.xyz/v1/read: $0.001 per call. Read a web page as clean Markdown for an LLM. Fetches the URL (robots.txt respected), extracts the main content (drops navigation, headers, footers, sidebars, cookie banners) and returns GitHub-flavoured Markdown with headings, lists, tables, code blocks (with language) and absolute links, plus metadata (title, description, author, published/modified dates, site name, language, canonical URL, image), an outline of headings, word count and a token estimate. Options: mode=full for the whole page, links=false to save tokens, images=true, include_links=true for the link list, max_tokens to truncate, chunk_tokens to get RAG-ready chunks with heading paths. Also converts native Markdown, plain text, JSON and RSS/Atom feeds. Never charged when it fails: unreachable or 4xx/5xx targets, robots.txt disallow, unsupported types (PDF, images) and empty/JavaScript-only pages return an error and are not settled. $0.001 per page.
  Example: GET https://readmd.canisell.xyz/v1/read?url=https%3A%2F%2Fexample.com%2F
- POST https://readmd.canisell.xyz/v1/read/batch: $0.005 per call. Read up to 10 web pages as clean Markdown in one call, in parallel: the same output as /v1/read for each URL, with a per-URL ok/error. $0.005 per call (= $0.0005 per page at 10 URLs). Charged only if at least one URL is read successfully.
  Example: POST https://readmd.canisell.xyz/v1/read/batch with JSON body {"urls":["https://example.com/","https://www.rfc-editor.org/rfc/rfc9309.html"],"links":false,"max_tokens":2000}

## Free endpoints
- GET https://readmd.canisell.xyz/v1/preview: Free preview of /v1/read for one URL (?url=): title, metadata, outline, word and token counts, and the first 500 characters of the Markdown. Rate-limited per IP.
- GET https://readmd.canisell.xyz/v1/sample: A recorded input and full paid response of /v1/read.
- GET https://readmd.canisell.xyz/openapi.json: OpenAPI 3.1 with prices (x-payment-info)
- GET https://readmd.canisell.xyz/.well-known/x402: x402 resource list
- GET https://readmd.canisell.xyz/v1/health: liveness

## MCP server (remote)
- URL: https://readmd.canisell.xyz/mcp (Streamable HTTP, stateless; no install, no API key). Client config: {"mcpServers": {"readmd": {"url": "https://readmd.canisell.xyz/mcp"}}}. Stdio-only clients: npx -y mcp-remote https://readmd.canisell.xyz/mcp
- Free tools: about, preview_url
- Paid tools (x402 MCP transport, _meta["x402/payment"]): read_url ($0.001), read_urls ($0.005). Without a payment they return the payment terms and the paid HTTP URL; nothing is charged.

## Notes
- Prices: read_url $0.001 per page; read_urls $0.005 per call for up to 10 pages (charged only if at least one succeeds).
- Never charged for failures: invalid URL (400), robots.txt disallow or unavailable, target 4xx, unsupported content type, empty or JavaScript-only page (422), target 5xx or unreachable (502), timeout (504).
- Limits: 3 MB per page, 5 redirects, 12 s per page, 400k characters of Markdown. No JavaScript rendering.
- robots.txt: product token `readmd` (falls back to the * group). User-Agent: readmd/1.0 (+https://readmd.canisell.xyz/llms.txt).
- Markdown is GitHub-flavoured; tokens_est is about characters/4 (1 per CJK character).