Whisper Keyboard

Whisper is an automatic speech recognition (ASR) system that unlocks the power of multilingual and multitask supervised data collected from the web. Using a simple end-to-end approach, the ASR is able to convert spoken words into text captions and has been trained on over 680,000 hours worth of data, making it one of the most robust models on the market.

This app takes Whisper's already impressive capabilities one step further. By introducing an easy-to-use 'keyboard' for Android, the user can simply tap the 'Record' button, and speak for up to 30 seconds, before tapping the 'Transcribe' button to convert their speech into text. The 'Newline' and 'Backspace' buttons make it easy to edit and fix transcription errors, while the user's favorite touch keyboard provides all other necessary features.

This app is restricted to English-to-English transcription only, but is still capable of recognizing a wide range of accents and background noise. It is ideal for those who prefer to use their voice to input text, but need that text to be accurate and understandable.