LLM Chat PDF. HTML & Markdown Exporter 作者： Fschwar4

Check: github.com/fschwar4/llm_copier for more details.

LLM Markdown Copier is a secure and lightweight Firefox extension that lets you export chat sessions from large language model (LLM) web apps as clean, shareable Markdown, portable PDF documents or raw HTML document object model (DOM). It currently supports OpenAI ChatGPT (chatgpt.com), Anthropic Claude (claude.ai) and Google Gemini (gemini.google.com). The extension is intentionally simple: everything happens entirely in your browser and nothing is ever sent to any server.

• One‑click export to Markdown. Convert the currently open chat into a structured Markdown document with a Quarto YAML header. The exported text is copied directly to your clipboard.
• Full HTML DOM capture. Need the raw page source for debugging or archival purposes? A third button allows you to copy the complete HTML document to your clipboard.
• High‑quality PDF downloads. Generate a PDF from your conversation using pdfmake, a client-side PDF generation library. No Canvas Approach! Leading to high quality pdf generation even for long conversations. The PDF is directly build based on the extracted Markdown content.
• PDF Layout Customization. You can customize the PDF output layout (font size, margin size, color scheme) in the extension settings page.
• Syntax Highlighting for code blocks. The PDF export includes syntax highlighting for code snippets using highlight.js. See below for a full list of supported languages.
• Minimal permissions. The extension uses the activeTab permission to execute scripts in the currently active tab only in response to a user action (clicking the extension button), the clipboardWrite permission to copy exported content to the clipboard, and the storage permission to save custom settings. It does not read from your clipboard or access unrelated browsing data. Host permissions are restricted to the ChatGPT, Claude, and Gemini domains only.
• Open source. Open source license (AGPL-3.0) You can inspect the code or build it yourself from the GitHub repository.

🔒 100% client‑side.
All data extraction and conversion happen in your browser — no requests to external servers, no telemetry, and no tracking.