PI Plamen Ignatov
Android app

Offline Transcriber

Private speech-to-text for your audio recordings. Offline Transcriber turns meetings, lectures, interviews and voice notes into text directly on your Android phone: the speech recognition runs on the device, so your audio is never uploaded to a cloud transcription service.

Record in the app or import an existing audio file, then read, search, summarise and export the result. No account is required.

Get it on Google Play ↗

Features

On-device transcription

AI speech recognition that runs locally. After the models are downloaded, transcription works without an internet connection.

AI summaries and chat

A local language model condenses long transcripts into the main points, and you can ask it questions about a recording.

Timestamps

Every line is timed, so you can jump to the moment something was said and play it back.

Search

Find a word or phrase across all your transcripts.

Export and share

Save or share transcripts as text, and with Pro as SRT and WebVTT subtitles or a Word document.

Multiple languages

Transcribe recordings in several languages, including English and Bulgarian.

Screenshots

List of recordings with a search field and a microphone button
Your recordings
Transcript with a timestamp for every line and an audio player
Timestamped transcript
Recording screen
Recording
Choosing speech recognition and summary models
On-device models
Welcome screen
Getting started

Free and Pro

Free

  • Recordings up to 10 minutes, transcribed on the device
  • The first 10 minutes of imported audio files
  • Summaries, translations and chat for recordings up to 10 minutes
  • Text export, with or without timestamps

Pro: a one-time purchase, no subscription

  • Recordings of any length and full transcription of imported files
  • Summaries and translations of recordings over 10 minutes
  • Export to SRT and WebVTT subtitles and to Word (DOCX)

Privacy

Recordings, transcripts and summaries stay on your device unless you choose to export or share them. There are no ads and no analytics, and I do not run a server that receives your audio. The internet is used only to download the AI models and for purchases through Google Play. The details are in the privacy policy (English and Bulgarian).

Good to know

Accuracy and speed depend on your device, the language, the recording quality and background noise. Larger models give better results but need more memory and time. AI summaries can contain mistakes, so check them when accuracy matters. Always get consent before recording a conversation.

Support

Found a bug or have a suggestion? Write to support@ignatov.eu.