Clarishot: Export ChatGPT, Claude, Gemini Chats as PDF, PNG, Markdown

Save ChatGPT, Claude and Gemini chats as PNG, searchable PDF or Markdown. Capture any web page, edit, record. Nothing is uploaded.

Clarishot turns your ChatGPT, Claude, and Gemini conversations into publish-ready screenshots, searchable PDFs, Markdown, JSON, and plain text. It also captures any other web page, lets you edit the picture in Studio, and records video. A camera button appears beside Copy on every message. Click it, and a pixel-perfect capture lands on your clipboard, ready to paste anywhere. The camera's menu also copies the message as rich text, Markdown, or JSON, or saves it as a PDF. Capture exactly what you need: - One message, even one that's ten screens tall. - A prompt + response pair with a single Option/Alt-click. - A hand-picked set or a range. Skipped stretches are labeled ("N messages omitted"), never hidden. - The entire conversation from the toolbar, in your choice of PDF, image, Markdown, JSON, or plain text: every message, in order, even ones scrolled far out of view. Long chats save as one clean multi-page PDF, or as a ZIP of page images if you asked for an image. Capture any web page (beta). On any other site, the Clarishot button captures the whole page from top to bottom as one picture, with cookie banners left out, or just the part you see. Alt+Shift+C (Option+Shift+C on a Mac) captures the whole page in one step. A very tall page saves as a PDF or a ZIP of images, cut between lines of text. Edit in Studio. A web page capture opens in Studio, and a chat capture can too. Blur or hide what others shouldn't see, or let "Hide private info" find e-mail addresses, phone numbers, keys, card numbers and more. Add arrows, shapes, text, numbered steps, callouts, and sticky notes, crop, and add a browser frame or a watermark. The page's own words stay in the picture as real text, so you can select, copy, and search them. Save as PNG or a searchable PDF, or copy the picture. Every PDF is searchable. Find, copy, and screen readers work in nearly every language: Latin, Greek, Cyrillic, Hebrew, Arabic, Chinese, Japanese, Korean, Hindi, Thai, and many more, plus math and symbols. PDFs come on A4 or Letter pages with page numbers and a bookmark for every prompt, or as one long page. Record what a screenshot can't show. Interactive artifacts, code previews, and Canvas apps are things you use, so Clarishot records them, and any other page too: pick a box or the whole tab and get an MP4, with every click shown as a ripple. Add the tab's sound, your voice, or your camera in a corner bubble, each off until you turn it on. The recording bar floats above your windows and never appears in the video. Built for the hard parts of AI chat interfaces: wide code blocks, tables, canvases, diagrams, and math render pixel-perfect, with real fonts and real syntax highlighting, exactly as the platform drew them. Clarishot doesn't rebuild your chat in a template of its own. It captures what was actually rendered. Private by design: - Runs in your browser. No servers of our own, no account, no analytics, no uploads. Text exports read your conversation from the AI platform you're signed in to, the way the page itself does; images and PDFs are composed from what's on screen. For a Google Doc, Studio reads the doc's own PDF from Google to give you its words. Nothing is sent to us or to anyone else. - Captures happen only when you click a Clarishot control, and recordings only when you start one. On other websites, Clarishot does nothing until you click. - Videos are recorded and saved on your computer. Your microphone and camera are used only while you record, and only if you turn them on. Free. No account, no quota. Privacy policy: https://clarishot.ai/privacy · Changelog: https://clarishot.ai/changelog · Support: support@clarishot.ai