Why Vox2Text exists
Voice messages are easy to record but not always easy to listen to: in a meeting, on a train, when you need to find something again or keep a record of a conversation. And for people who are deaf or hard of hearing, text is essential. Vox2Text turns them into text you can read, copy and keep.
How it works
Vox2Text uses Whisper, an open-source speech recognition model, running right in your browser. Audio is processed on your device and never uploaded to a server: that’s why there’s no sign-up and nothing to install.
Really free
No subscriptions, no ads, no data collection. The service is supported by voluntary donations: if you find it useful, you can help keep it running, or simply recommend it to someone.
Who’s behind it
Vox2Text is an independent project started in Italy in 2026. It is not affiliated with WhatsApp or Meta.
Contact
Suggestions, bug reports or partnerships: giuseppe.cuccurullo@gmail.com
Credits and licenses
Vox2Text is built with free software:
- Whisper (OpenAI) – MIT
- Transformers.js (Hugging Face) – Apache 2.0
- ONNX Runtime (Microsoft) – MIT
- Nunito – SIL Open Font License
- Lucide – ISC