Analysis updated 2026-05-18
Record meeting notes or interviews on your iPhone with automatic real-time transcription saved alongside the audio file.
Import existing voice memo audio files and transcribe them on-device in one of six supported languages.
Add on-device AI-generated summaries and topic tags to long recorded meetings without reading the full transcript.
| iamwilliamli/livetranscriber | asaptf/swift-language-models | crafcat7/peakmon | |
|---|---|---|---|
| Stars | 6 | 6 | 7 |
| Language | Swift | Swift | Swift |
| Setup difficulty | hard | moderate | easy |
| Complexity | 3/5 | 4/5 | 3/5 |
| Audience | developer | researcher | developer |
Figures from each repo's GitHub metadata at analysis time.
Requires Xcode beta with the iOS 27 SDK and an iOS 26 or later device or simulator, FoundationModels availability depends on the device.
LiveTranscriber is an iOS recording and transcription app that works entirely on your device. You tap to record, and the app transcribes your speech in real time as you speak, using Apple's built-in Speech APIs. No audio or transcript text is sent to an external server. Recordings are saved locally by default, with an option to sync them across your Apple devices through iCloud. The app captures stereo audio and outputs WAV or M4A files. Transcripts are timestamped line by line, so you can tap on any line to jump to that point in the recording during playback. After stopping a recording, you give it a name, add tags, and optionally attach a location. The app then uses Apple's on-device language model to generate a short summary and topic tags for the saved transcript. While recording, LiveTranscriber shows a Live Activity on the Lock Screen and in the Dynamic Island, displaying elapsed time, the latest transcribed text, and a stop button. A Home Screen widget provides quick access to start recording, open saved files, or go to settings. The file library lets you search by filename, transcript content, language, or tags. You can also import existing audio files from the Files app or the iOS share sheet and transcribe them after the fact, including re-transcribing in a different language. The app supports English, Simplified Chinese, Traditional Chinese, German, Dutch, and Japanese. Building it requires Xcode with the iOS 27 SDK, which was in beta at the time this was published. The source is available under a custom license: personal and non-commercial forks are permitted, but commercial products built on it require visible in-app attribution to the original project. White-label or attribution-free commercial use requires separate written permission from the author.
An iOS app that records audio and transcribes speech in real time on-device using Apple Speech APIs, with Lock Screen Live Activities, iCloud sync, and on-device AI summaries.
Mainly Swift. The stack also includes Swift, SwiftUI, iOS.
Free to fork and learn from for personal or non-commercial use, commercial products must include visible in-app attribution, and white-label commercial use requires written permission.
Setup difficulty is rated hard, with roughly 1h+ to a first successful run.
Mainly developer.
This repo across BitVibe Labs
Verify against the repo before relying on details.