Skip to content
Mobile Byte Sensei

Awaazly

Private transcription on the phone: Whisper speech recognition and summaries, in twelve languages.

Get it on Google Play (opens in a new tab)Visit website (opens in a new tab)
Status
Live
Category
Voice
Platforms
  • Android
  • AI
Home screen with a large record button, audio file import and recent transcripts

Transcription and summaries on the phone.

Meetings, interviews and lectures are hard to capture, and most transcription tools ask for a trade-off: upload the audio to the cloud, or accept weaker speed, language coverage or polish.

Awaazly transcribes live speech or an imported audio file on the phone with Whisper, and a local language model writes summaries, extracts action items and answers questions about the transcript. There is no account and no upload.

It is for journalists, founders, students, researchers and bilingual creators whose recordings need to stay private.

Who it is for.

  • Journalists and podcasters

    Transcribing confidential interviews without sending audio to a third party.

  • Founders and sales leads

    Capturing meeting decisions and action items.

  • Bilingual creators

    Transcribing Hindi and English content.

  • Students and researchers

    Transcribing lectures without a connection.

What it does.

  • Transcription on the device

    Whisper speech recognition runs on the phone, with no server round trip.

  • Live recording

    Record with a live waveform, pause and resume, then save to the library.

  • Import an audio file

    Open an existing MP3, M4A, WAV, OGG, FLAC or AAC file and get a searchable transcript.

  • Twelve languages

    English, Spanish, French, German, Portuguese, Italian, Dutch, Russian, Chinese, Japanese, Korean and Hindi, chosen per recording or detected automatically.

  • Summary, action items and questions

    A local language model writes a short summary, extracts action items and answers questions about a transcript.

  • A searchable library

    Keyword search with highlights, starring and archiving, and filters by date, duration or source.

  • Edit in place

    Correct a transcript directly, with changes saved as you go.

  • Export and share

    Export text or captions through the share sheet. Only what you choose leaves the device.

  • Model and network control

    Choose the Whisper Tiny, Base or Small model, and allow downloads on Wi-Fi only, on mobile data, manually or never.

How it works.

  1. Set up

    Grant microphone access, pick a language and download a Whisper model, or skip the download for now.

  2. Capture

    Record live, or import an existing audio file.

  3. Transcribe on the phone

    Whisper and a local language model process the audio on the device, and the transcript is saved to the library.

  4. Work with the transcript

    Read, edit, summarise, extract action items or ask questions, then export or share only what you choose.

Screens.

Screens from the product design.

  • Live recording with a waveform, running timer, live caption preview and pause and finish controls
  • Transcript of a team standup with its duration, language, the Whisper model used and the full text
  • On-device AI summary of a meeting transcript, with copy, share and save actions
  • A question asked of a transcript, answered on the device, about the blockers raised in a meeting
  • Privacy settings: on-device processing, local recordings, encrypted transcripts and opt-in crash reports

Privacy and data.

Audio, transcripts and summaries are processed and stored on the device. Nothing leaves it unless you export or share it.

  • No account is required.
  • The privacy policy states that no analytics, tracking, crash-reporting or advertising SDKs are used.
  • Network access is used to download models from a public model host and for sharing you start yourself.
  • Permissions: the microphone, and optional file access for importing audio.

Availability and specification.

Status
Live
Availability

Includes applied AI.

Resources
Category
Voice
Business model

Free to use, with no subscription, in-app purchases or ads. An optional Pro tier is planned, with the free tier intended to remain complete.

Built with
  • Kotlin Multiplatform
  • Compose Multiplatform
  • whisper.cpp
  • llama.cpp
  • Llamatik
  • Room

Questions about Awaazly.

Does my audio ever leave my device?

No. Recognition and summaries run on the device, and audio and transcripts stay there unless you export or share them.

Do I need an account or a subscription?

Neither. Awaazly is free and needs no sign-up.

Which languages are supported?

English, Spanish, French, German, Portuguese, Italian, Dutch, Russian, Chinese, Japanese, Korean and Hindi, chosen per recording or detected automatically.

Can I transcribe an existing recording?

Yes. Import the audio file to get a searchable, editable transcript.

How is it private if it uses AI?

The models run on the phone, and Settings lets you restrict model downloads to Wi-Fi, allow mobile data, download manually or never.

More apps.

All products

  • Betterflo

    Voice

    Live

    On-device dictation in eleven Indian languages, in any Android app. No account.

    • Android
    • AI

  • PDF Tools

    Productivity

    Live

    Merge, split, compress, sign and convert PDFs on the phone, with no upload and no account.

    • Android

  • Status Saver

    Utility

    Live

    Save the statuses you have viewed, convert video to MP3 and compress media, all processed on the phone.

    • Android

Get Awaazly.

Private transcription on the phone: Whisper speech recognition and summaries, in twelve languages.

Get it on Google Play (opens in a new tab)Visit website (opens in a new tab)

Mobile Byte Sensei also builds for clients. View services