FluidScribe

FluidScribe turns your voice into precise text – entirely on your iPhone. No cloud, no account, no tracking. Just start talking: the text appears live on screen while you record. When you stop, FluidScribe transcribes the whole recording, gives it a fitting title and sums it up with action items. PRIVATE BY DESIGN • Speech recognition, titles, summaries and translations run on your device • Your recordings and texts never leave your iPhone • No account, no ads, no analytics • Works in airplane mode once the speech model is downloaded PRECISE RECOGNITION • NVIDIA Parakeet, one of the most accurate open speech recognition models, optimized for the Neural Engine • 25 European languages, even mixed within one memo • Punctuation and capitalization included • An hour of audio transcribed in seconds MORE THAN A TRANSCRIPT • Live transcript while you record, with pause and resume • Titles, summaries and checkable action items, created by Apple Intelligence • Speaker detection for conversations and meetings, with your own names • Tap a word and playback starts right there • On-device translation, even for memos that mix languages • Ask your memos: answers from everything you have recorded, with sources • Search, favorites and six sort orders RECORD WHEREVER YOU ARE • Apple Watch: record on your wrist, your iPhone transcribes in the background and lets you know when the text is ready • Control Center, Lock Screen, Action button, Siri and Shortcuts • Live Activity with pause and stop • Import audio from Voice Memos or Files • Share text, audio or both GOOD TO KNOW On first launch, FluidScribe downloads the speech model once (about 600 MB, the compact model about 220 MB). Titles, summaries and Ask Your Memos require a device with Apple Intelligence; without it, memos are named by date and time. Privacy Policy: https://fluidscribe.weatherguide.app/en/privacy/ Terms of Use: https://www.apple.com/legal/internet-services/itunes/dev/stdeula