Add audio or video
Choose a media file supported by macOS, then set a clear title and date. The original stays unchanged.
Local speech recognition on macOS
Convert audio and video to text on your Mac
Import a call, lecture or interview. AnyBrief transcribes it locally and, when speaker separation is enabled, organizes voices into readable turns.
For macOS 14 or newer, Apple Silicon and Intel. Developer ID signed and Apple notarized.
From a recording to a working transcript
No browser upload and no manual handoff between transcription and note-taking tools.
Transcribe locally
Use FluidAudio or whisper.cpp. Language options depend on the engine; speaker separation is configured separately.
Read, verify and export
The transcript and a processed audio copy stay in the meeting; a summary is created separately when configured.
What happens to the file
Import does not change or move the original.
- Audio extraction
- AnyBrief reads the first available audio track. If macOS cannot decode one, import stops and no meeting is created.
- Recognition copy
- The app creates a separate local mono PCM WAV at 16 kHz. For multitrack media, only the first readable audio track is imported.
- Meeting files
- AnyBrief keeps
transcript_raw.txt, workingtranscript.txt, optionalsummary.md, and a local archive containing processed audio and technical data. - Reprocessing
- You can repeat transcription with another model. The raw transcript remains separate from the LLM-cleaned working version.
More than a raw transcript
Clean the text or draft a brief, minutes or structured notes with your chosen LLM and prompt.
- Speaker-separated turns when diarization is enabled
- Repeat transcription with another model
- Custom prompts for summaries and cleanup
- Transcript and summary export to an existing folder
A summary needs a configured prompt and LLM and is created for transcripts with at least 30 recognized words. Verify ASR and LLM output against the recording.
Common questions
Which audio and video formats can AnyBrief transcribe?
Common examples include M4A/AAC, MP3, WAV, AIFF, FLAC, OGG and CAF, plus MOV, MP4, M4V, AVI, MPEG and 3GP files with an audio track. AnyBrief's automated tests cover M4A/AAC, WAV and MOV. Import support depends on the codec inside the file and the macOS version; a file without a readable audio track is rejected.
Is transcription really local?
Yes. FluidAudio and whisper.cpp run on your Mac. For summaries, choose local Ollama or a remote provider you configure.
Can it record a live call too?
Yes. AnyBrief records system audio and microphone audio into separate tracks.
Related workflows
Turn your first recording into text
Install AnyBrief and import an audio or video file.