SendPaper vs FrankenWhisper

Sendpaper puts a button in Safari that sends whatever you're reading straight to your reMarkable — no cables, no email-to-yourself, no copy and paste. Hit the toolbar button on any article and it arrives as a clean, reflowable EPUB: just the text and images, with the ads, sidebars, popups, and clutter stripped away. It's the reading experience e-ink was made for. A SMARTER READER Sendpaper recognises the site you're on and extracts exactly the right content, so pages that never used to convert well now come through beautifully: • AI chats — send a ChatGPT, Claude, or Gemini conversation and read the whole thread as a document, every prompt and reply in order. • YouTube — send a video and get its full transcript as an article, complete with the title, channel, and chapter headings. • Discussions — Reddit threads and Hacker News posts arrive readable and structured. • Reference & news — Wikipedia, Substack, and major news sites convert faithfully, with proper headings and images. • Everything else — any ordinary article is cleaned up and reflowed automatically. PRIVATE BY DESIGN Pages are converted right on your device and sent straight to your own reMarkable account. There's no Sendpaper server in the middle reading your pages, and no account to create with us. HOW IT WORKS 1. Pair once with your reMarkable using a one-time code (the same secure device-pairing the official apps use). 2. Open any article, tap the Sendpaper button, and it syncs to your reMarkable. 3. Pick it up on your device and read. Works on iPhone, iPad, and Mac in Safari.

Turn recordings and videos into useful text without uploading them. FrankenWhisper transcribes microphone recordings, audio files, and playable videos from Photos, Files, or the share sheet. The full-quality lane uses the large-v3-turbo model. Live dictation and the optional custom keyboard use a smaller model for low latency, with local noise reduction in the real-time path. Speaker identification can separate up to four voices. Add names when you know them or leave them blank and use color alone. Export the transcript as text, Markdown, HTML, SRT, or JSON with timestamps and speaker structure. Subtitle Studio creates a new H.264 video with customizable captions, font, size, position, and color. Word-level timing drives a karaoke highlight, and named speakers can receive distinct labels and color schemes. All transcription, speaker analysis, denoising, timing, and video rendering run on the device. Model downloads are the only required network activity. The keyboard's Full Access setting is used only for the local App Group handoff between the keyboard and FrankenWhisper; the keyboard extension has no network client.