Voice Notes - Offline to Text

Record, and it turns into text on your device.

Voice Notes turns lectures, meetings and interviews into text with on-device Whisper speech recognition. Your audio, transcripts and summaries are never sent to a server. (The internet is used only for the first speech-model download.)

[Features]
• Near real-time transcription — text fills the screen as you speak
• Works offline — after a one-time model download, it runs in airplane mode
• Privacy first — no account, no ads, no tracking, no recordings or text sent out
• File transcription — MP3, WAV, M4A, WEBM and FLAC files become text on your device
• Refine — re-transcribe a saved recording with a larger model for better quality
• Key sentences — pick out the important sentences from a long recording
• Multilingual — Korean, English, Japanese and 20+ languages, with auto-detect
• Export — TXT, Markdown, SRT subtitles, clipboard, or share the original audio
• Read along — the paragraph you are hearing is highlighted during playback

Made for people who record meetings, lectures and interviews, and for anyone whose recordings must not leave the device, such as medical or legal work.

[Good to know]
• A speech model (about 75MB for the lightweight one) must be downloaded once before you can record
• Keep the app open while recording. Locking the screen or switching apps pauses recording
• Accuracy depends on the recording environment, accent and the model you choose, so please check important content against the audio