Skip to content

Company

About Vermin

Who builds Vermin, why it exists, and how the company behind it works.

Vermin is a web data API for AI applications. Give it a URL and it returns clean, LLM-ready markdown, token-counted chunks and metadata. Give it a site and it maps or crawls it. Give it a domain and it returns the company's logo, colours and fonts. It is built for developers and AI agents who need the content of the web without maintaining scrapers, headless browsers and boilerplate filters themselves.

Why it exists#

Vermin started as the scraper inside PageClerk, a product that trains customer-support chatbots on a company's own website. Chatbots are only as good as the context they are given, and most of a typical web page is navigation, cookie banners, footers and repeated links. We built an extractor that removes that noise, escalates from a plain HTTP fetch to a headless browser only when a page needs it, and reports exactly what each page cost. On our own 852-page benchmark it raised metadata coverage from 49% to 99% and brand-name accuracy from 64% to 93%, at the same cost per page. That extractor became Vermin.

How we work#

  • Pay for results, not attempts. One scrape is one credit, browser rendering included. If a page is blocked, times out or returns nothing usable, it costs 0 credits and you get a typed error code.
  • Good manners. Scrapes and crawls respect robots.txt and identify themselves.
  • Built for agents. A hosted MCP server, an OpenAPI 3.1 spec, llms.txt and Markdown versions of our pages, so an AI agent can set Vermin up without a human filling in forms.
  • Edge-native. The API runs on Cloudflare's network, close to the sites it reads.

The company#

Vermin is a product of Reshapify, LLC, a Delaware limited liability company.

Postal address: Reshapify, LLC, 1111B S Governors Ave STE 23005, Dover, DE 19904, United States.