LiteRec / Practical guide
Transcribe audio and video locally on your Mac
Updated October 13, 2026 · Based on LiteRec 1.2.4
A transcript makes a long recording easier to revisit. LiteRec can process audio locally after you download a speech model.
Set up once
- Open History & Transcripts, or press Command-T.
- Import an audio/video file, or choose Transcribe for a saved recording.
- Select local processing and download a speech model. Model downloads use the network; local transcription does not upload the recording.
- Choose the language if needed and start transcription.
Recording requires macOS 13+. The Base and speaker-labeling models support macOS 13+, while the bundled Small and Large presets require macOS 15+. Inference has been tested on Apple Silicon; performance on older hardware varies. Download size is shown before setup.
Review before you reuse
Speech recognition can mishear names, figures, accents and overlapping voices. Replay passages, correct the text and adjust speaker labels where necessary. Labels group voices within the recording; they do not identify people automatically.
Export for the next step
Use TXT or Markdown for readable notes, SRT or VTT for subtitles, and JSON for a transcript backup. The original media must remain available for playback. Moving a file can require locating it again.
Local and cloud modes have different boundaries
After purchase, local transcription has no additional usage fee; each file is limited to five hours. Optional cloud transcription sends extracted audio to your selected provider after confirmation. It uses your own API key and billing account, and the current extracted-audio upload limit is 24 MB. A one-time app purchase does not include unlimited third-party API usage.
Start with a short representative file. For private content, stay in local mode and consider whether your chosen save folder syncs to another cloud service.