url-to-markdown-api

原文涵盖的主题

• URL to Markdown - Web Page Extract API
• When to trigger
• Setup
• If you do not have a key yet
• Endpoint
• Cost is a function of the body
• Parameters
• What the three formats actually are

Read any web page and get it back as clean Markdown, plain text, or raw HTML. This is the read-a-page primitive: one URL in, page content out. Extract is not a platform - it is a core endpoint. There is no namespace, no per-site parser and no site-specific parameters. It works on any http(s) URL. Use this skill when the user asks to: Read, summarise, or quote a specific web page Turn a URL into Markdown or plain text for an LLM prompt or a RAG chunk Pull an article, docs page, changelog, pricing page or blog post into the conversation Fetch a page that blocked a plain HTTP request Grab the raw HTML of a page to parse it yourself This is usually the right first tool whenever a user pastes a link and asks a question about what is on it. Many pages from one site, or a long list of URLs? Do not loop this endpoint. Use the website-crawler-api skill: POST /api/v1/extract/map lists a site's URLs from its sitemap for 1 credit, POST /api/v1/extract/crawl crawls a whole site to Markdown as one job, and POST /api/v1/extract/batch extracts up to 1,000 URLs as one job, both billed per successful page at this same tier price. See Website Crawler API. Get a free API key at scavio.dev (50 free cre

这里只展示节选，请以完整原文和当前要求为准。