Compare

Viaalink Notetaker vs Real-Time Voice Cloning

Real-Time Voice Cloning is not a meeting notetaker but a research project (a master's thesis): it turns a few seconds of audio into a digital voice that speaks any text. Notetaker does the reverse: it turns spoken words into text.

Where Notetaker is stronger than Real-Time Voice Cloning

  • Phone and computer

    Android, and Windows (alpha), macOS and Linux. The phone sends a session to your computer over your own Wi-Fi.[2]

  • No bot in your call

    You start Notetaker yourself on your phone or computer. Nobody extra joins.[1]

  • Everything stays on your device

    Speech becomes text on your own phone or computer. Audio, transcripts and reports never reach Viaalink.[2]

  • 99 languages

    On your phone Notetaker recognises speech in all 99 languages Whisper knows. We measured the accuracy for Dutch and English.[3]

  • Transcribe existing recordings

    Also audio and video files you already have, on your phone and on your computer.[4]

  • Voice profiles with consent

    Can also be made afterwards from a recording, and they stay on the device.[5]

Side by side

Only what works today. Where their site is silent, we say so.

Viaalink Notetaker and Real-Time Voice Cloning side by side
TopicViaalink NotetakerReal-Time Voice Cloning
A bot in your callNo bot You start Notetaker yourself on your phone or computer.[1] Not stated on their site[9]
Where speech becomes textOn your device On your own phone or computer.[2] Not stated on their site[9]
Where recordings and text are keptOn your device Audio, transcripts and reports never reach Viaalink.[2] Not stated on their site[9]
What writes the reportYour phone A language model on the phone itself. The report never reaches Viaalink.[2] Not stated on their site[9]
Phone and computerBoth, together Android, and Windows (alpha), macOS and Linux. The phone sends a session to your computer over your own Wi-Fi.[2] Notetaker is ahead hereComputer only Windows and Linux, as a Python project.[9]
Without internetYes After activation; internet only to fetch the speech model once, to activate and renew your license, and, if you agree, to check for updates.[6] Not stated on their site[9]
Languages99 languages Phone and computer: all 99 Whisper languages. Measured for Dutch and English.[3] Not stated on their site[9]
Voices and speakersOn the device Only with consent; on your phone also afterwards from a recording. Recognition with a profile is in beta, separating voices without one in alpha.[5] Voice cloning A voice from a few seconds of audio.[9]
Existing recordingsYes Audio and video files, on phone and computer.[4] Not stated on their site[9]
ExportYes Transcript and report, as PDF, Markdown or Word.[7] Not stated on their site[9]
Account neededFor the license The activation code is on your account page.[8] Not stated on their site[9]

Coming soon: reports on the computer with AI models on your own machine, your own key for an AI service, and connecting Claude, ChatGPT or Mistral to your phone and computer.

Where Real-Time Voice Cloning is stronger

  • It can recreate a voice; Notetaker does not generate speech.[9]
  • The code is public on GitHub.[9]

Prefer online? Viaalink Social

Viaalink Social is the phone app for meetings, chat and sharing, online too. An AI agent can work with it through MCP. Social is coming to Android.

See Viaalink Social

Try Notetaker yourself

Notetaker is in beta. See for yourself how it works in your own conversations.

During the beta period we are still analysing our pricing.