Scrieb is local transcription software for Windows and macOS. Speech-to-text, speaker identification, and summaries all run on the user’s own computer using OpenAI’s open-source Whisper model, so recordings, transcripts, and summaries are never uploaded to a server. After the initial model download, Scrieb works fully offline.

Users can drop in audio or video files (MP3, WAV, M4A, MP4, and more) or record directly, including Zoom, Teams, and Google Meet calls via system audio—with no bot joining the meeting. The transcript identifies who said what, includes timestamps, and comes with a locally generated summary. Results can be exported to DOCX, PDF, or TXT. Scrieb supports 90+ languages, has no per-minute limits, and lets users choose the Whisper model size in Settings.

Because the content never reaches a third-party vendor, Scrieb is used by lawyers, therapists, clinicians, researchers, and journalists whose recordings cannot be sent to a cloud service. For the content processed locally, there is no data processing agreement or business associate to sign, and no cross-border data transfer. The data flow is documented at

Scrieb is free to try, and one licence covers both Windows and Mac.