LLM Copy Buttons + RAG Library

Extracts main content and prepares Markdown/Plain text for LLMs on demand.

LLM Copy Buttons helps turn web pages into clean, useful content for AI models. ✅ What problem does the extension solve: When an LLM reads a website, it sees a lot of noise: menus, ads, buttons, footers, and other clutter. As a result, the model loses context, gets distracted by irrelevant blocks, and may answer inaccurately. The extension lets you extract clean, structured information from any page in one click. Clean input means better output. 🆓 What’s included in the free version: Copy for LLM Copies a page in an AI-friendly format: title, URL, key metadata, and clean text. View Markdown Shows a structured version of the text: headings, lists, code blocks, and tables. View Plain Text Shows plain text without formatting. Open in ChatGPT / Claude / Perplexity Quickly opens the selected service in a new tab and puts it into context. Save to Library Saves the page to your local library. Library A convenient list of saved materials: search, select, delete, export. 💎 What the Pro version adds: Full RAG Library Save articles, documentation, READMEs, and other sources in Library, then work with them as one unified knowledge set. Bulk extraction from a URL list Paste a list of links into Extract URL list, and the extension processes them one by one. This is useful for quickly building a dataset. Export a ready-to-use RAG package (ZIP) Pro lets you export your library without limits in a single archive. Inside, you get everything needed for downstream processing: 📝 markdown/ — structured page texts (.md) 📄 plain/ — plain text (.txt) 🧩 meta/ — source metadata (.json) 🧮 pages.jsonl — pages in pipeline-ready format 🧠 chunks.jsonl — vectorization-ready segments ⚙️ manifest.json — export package description Smart Chunking For RAG, it is important not only to split text, but to preserve meaning across boundaries. Library settings include: chunk size (number of characters) overlap (chunk overlap) This helps produce more stable retrieval and more accurate LLM answers. 👥 Who Pro is for people building an internal knowledge base teams preparing content for AI assistants users who need regular export of large datasets developers building RAG solutions 🔒 Privacy and security works only on explicit user actions no hidden background tracking no page content sent to third-party servers minimal browser permissions