How do I use AI Transcribe in Noted?
Last updated
Open a note with a recording, tap ✨, choose Transcribe Audio, check the language and tap Start. It runs on your device, can also transcribe live while you record, and you can insert the transcript into your note or export it as a PDF. EditorShort answer
Open a note with a recording, tap ✨, choose Transcribe Audio, check the language and tap Start. It runs on your device, can also transcribe live while you record, and you can insert the transcript into your note or export it as a PDF.
What is AI Transcribe in Noted?
AI Transcribe turns the audio in a note into a written transcript, with a timestamp on every line. Tap any line to hear that moment in the recording. It runs entirely on your device: on iOS 26, iPadOS 26 and macOS 26 it uses Apple’s built-in speech recognition, and for languages Apple doesn’t support yet, or on older systems, it uses a Whisper model that you download once.
How do I transcribe a recording?
- Record in a note, or import an audio or video file (⋯ → Import).
- Tap ✨ on the toolbar and choose Transcribe Audio.
- Check the Language. To label who is speaking, turn on Identify speakers.
- Tap Start. Lines appear as they’re transcribed; keep Noted open until it finishes.
- When it’s done, tap any line to jump to that point in the audio, tap 🔍 to search the transcript, or tap ⋯ and choose Insert to note or Export as PDF.
If Noted uses Whisper for your language, it first asks you to choose a Model and download it. Larger models are more accurate but take more space and run more slowly.
On a Mac, open the note’s AI menu and choose Transcribe audio.
Does AI Transcribe work while recording?
Yes. While a recording is running, tap ✨ and choose Transcribe Audio. To have every recording transcribed automatically, go to Settings → Transcribe and set Transcribe while recording to Always. See How do I use live transcription?
Can it tell who is speaking?
Yes. Turn on Identify speakers before you tap Start, and each line is labelled with a speaker. Touch and hold a line and choose Assign Speaker to give a speaker a name. This needs a language supported by Apple’s speech recognition; with a Whisper model the option isn’t shown.
How accurate is AI Transcribe in Noted?
AI Transcribe is designed to quickly convert audio recordings to text, and is at its most accurate in clear, quiet environments. Background noise, speaker clarity and complex language can all affect the result, and accuracy may vary by language. Our team is constantly working on improvements to make transcription more reliable across different environments.
Is my audio sent anywhere?
No. Transcription runs on your device, so your recordings and transcripts stay private.
How long does it take?
It depends on the length of the recording, your device and the model used. Most recordings finish within a few minutes; larger Whisper models take longer.
Is there a limit on how much I can transcribe?
No. There’s no cap on audio length, number of recordings or words.
Which languages does it support?
Most major languages are covered by Apple’s speech recognition. For other languages Noted uses Whisper, which supports many more. Accuracy can vary by language.