Sagt turns audio recordings into text – entirely on your own iPhone or iPad. Record a meeting, or share a recording from Voice Memos or Files, and get an accurate transcript with timecodes and who said what, ready to search, edit and share. Ask every interview in a project, and play the quote the answer is based on.
LOCAL AND PRIVATE
• Your audio and text never leave your device. No account, no cloud.
• The language model is downloaded once. After that, recording and transcription work offline.
RECORD MEETINGS – EVEN WITH THE SCREEN LOCKED
• Start recording, put your phone on the table and lock the screen. Sagt keeps recording.
• When you stop, the recording is transcribed automatically, and the work can continue in the background.
SHARE FROM OTHER APPS
• Share audio files and videos from Voice Memos, Files and other apps straight to Sagt.
• Your recordings and transcripts are also in the Sagt folder in the Files app.
FREE – WITH SAGT PRO AS ONE PURCHASE
• Free: every model, including NB-Whisper, for recordings of up to 10 minutes. Edit, get summaries, copy and share the text.
• Sagt Pro is a one-time purchase, not a subscription, and works on Mac, iPhone and iPad: long recordings, who said what, projects, search across all recordings, export to Word, Markdown and subtitles, timecodes and audio clips.
MADE FOR NORWEGIAN – AND ENGLISH
• For Norwegian: NB-Whisper from the National Library of Norway, trained on Norwegian speech, also on iPhone.
• Sagt is available in Norwegian, English, Swedish, Danish and German. Swedish and Danish get their own models: KB-Whisper from the National Library of Sweden, and a Danish Whisper Turbo trained further on about 12,900 hours of Danish speech. German uses Whisper large-v3 and large-v3 turbo from OpenAI, which are very good at German.
• Choose between several models, from small and fast to most accurate. Sagt recommends the one that suits your iPhone or iPad.
• Transcribe again with another model and compare the versions.
• Add your own names and terms for more accurate spelling.
WHO SAID WHAT (PRO)
• Sagt finds the speakers while the recording is transcribed.
• Name the speakers, and their names appear right in the text.
EVERYTHING AROUND YOUR RECORDINGS
• Add several recordings. Sagt works through the queue and shows the text as it comes in.
• Play back at any speed, with the sentence being read highlighted. Tap the text to jump in the audio.
• Search all your recordings and jump straight to the moment (Pro).
• Group recordings into projects (Pro).
• Copy and share the text, or save it as plain text. With Pro: Word, Markdown, SRT and WebVTT, with timecodes.
ASK YOUR RECORDINGS – AND HEAR THE ANSWER YOURSELF*
• Ask a question about every interview in a project (Pro) or about one recording: “What did they say about the budget?”
• The answer shows who said it (speakers, Pro) and in which recording, with the time and the excerpt it is based on.
• Play the quote with one tap and hear that it was really said – or open the recording right at that moment.
SUMMARIES AND ACTION ITEMS*
• A short summary with decisions and action items with an owner and a deadline.
• The timecode next to each action item plays that moment in the recording.
• Copy the summary with one tap and paste it into Jira, Linear or GitHub.
• Automatic titles. Everything runs on your device with Apple's language model.
*Requires an iPhone or iPad that supports Apple Intelligence, with Apple Intelligence turned on. Transcription works without it.
REQUIREMENTS
• iOS 26 or iPadOS 26
• iPhone with A15 Bionic or later: iPhone 13 or later, or iPhone SE (3rd generation)
• 3–8 GB of free space, depending on the model, and internet the first time to download the language model
Show more
