ScrapLLM - Web to Markdown for AI — dekrezz

ScrapLLM converts whatever is on screen into Markdown that is ready to paste into an LLM.

<b>What it copies</b>
<ul><li><b>Convert & Copy</b> — the page as Markdown; the article body by default, the full page on demand.</li><li><b>Copy Selection</b> — only what you highlighted, prefixed with the line range and source. Code is detected and fenced with its language.</li><li><b>Copy Chat</b> — an LLM conversation with roles and timestamps, the whole thread or the last N exchanges.</li><li><b>Download Markdown</b> — the same output as a .md file.</li><li><b>Multi-tab</b> — several selected tabs at once, merged into one file or zipped separately.</li><li><b>Research</b> — one question in, one Markdown file out: the extension searches the web, reads each source, converts it and downloads the merged document.</li></ul>

<b>Site-aware extraction</b>
Reddit posts with their full comment tree, X threads with replies and quoted posts, and chat transcripts from ChatGPT, Claude, Gemini, Grok, Perplexity, DeepSeek, Copilot, Cursor and self-hosted front-ends.

<b>Privacy</b>
Page content is converted locally and is never sent anywhere. The optional token counter fetches static tokenizer data from tiktoken.pages.dev once and caches it. The Research feature is the one part that talks to a third party: it sends your question to DuckDuckGo to find sources, then fetches each accepted source itself, without your cookies. Everything else works offline.

<b>Open source</b>
MIT licensed. Source: https://github.com/dekrezz/ScrapLLM