Whisply
"Local audio transcription, native and offline."
Whisply runs OpenAI's Whisper model entirely on your Mac via whisper.cpp — every byte of audio stays on your device. Drag any .mp3, .wav, .m4a, .flac, .ogg, or video file onto the window and Whisply transcribes it locally; play back from any sentence with click-to-seek; export the result as plain text, Markdown with timestamps, or SRT subtitles. There's a built-in recorder for the microphone and a separate one for meeting audio (Zoom, Meet, Teams) via ScreenCaptureKit — useful when you want the transcript of a call without trusting anyone else with the recording.
What's in it
- On-device transcription via whisper.cpp — no cloud round-trip, no upload, no API key.
- Drag-and-drop or Cmd+O to import; multi-file queue processes in order with cancel + per-file progress.
- Built-in recorder: microphone, or system audio (Zoom / Google Meet / Teams) via ScreenCaptureKit.
- Inline audio player with click-to-seek timestamps and the playing segment highlighted as you go.
- Export to plain text, Markdown with timestamps, or SRT subtitles. Cmd+E from any transcript.
- Video support too — drop a .mp4, .mov, .mkv, .avi, or .webm and Whisply pulls the audio out (ffmpeg helps if installed).
- SwiftUI + SwiftData, signed for the Mac App Store sandbox. No accounts, no analytics, no telemetry.
On the roadmap
- Speaker diarisation — automatically detect who said what and label segments with Speaker 1 / Speaker 2 / etc.
- Per-speaker filtering — jump straight to one speaker's lines, or export a single speaker's transcript on its own.
Whisply hasn't shipped yet — we're targeting TestFlight first, then the App Store and Google Play. Drop us a line if you want to be on the early build list.
Get on the listNo build yet — but the engine is real.
Whisply is in active development. Screenshots and a TestFlight link will land here once we cut the first build.