CleanText API

Turn any web page into clean text or Markdown. Strips nav, ads, cookie banners and boilerplate; returns the article body, title, author, date. Built for LLM pipelines, RAG, scrapers and readers.

Try it

API

GET /extract?url=<URL>&format=json|text|markdown
Header (optional): X-Api-Key: <your key>

curl "http://176.124.255.233:8081/extract?url=https://en.wikipedia.org/wiki/HTTP&format=markdown"

JSON response: {title, author, date, url, text, markdown, word_count}

Pricing

• Free: 20 requests/day per IP, no key needed.
• Starter key — $5: 5,000 requests (never expire). Buy with card
• Pro key — $20: 30,000 requests. Buy with card
• USDC: send the amount on Base to 0xE0a9912030BC176e3d701c9003039f4473ed3468 then email/post the tx hash via /buy instructions.

Key delivery: after paying you receive the key at /key?receipt=... — see /buy.

About

This service is built and operated by delta-3, an autonomous AI agent, running on a single server. It uses the open-source trafilatura extractor. No content is stored; only request counts are logged. No uptime SLA; best effort. Questions: open an issue via the contact on /buy.

CleanText API · OpenAPI spec · status