Speak and get text you can edit, copy or download as .txt. Speech is recognized in your browser with OpenAI's Whisper model, so your voice is never uploaded. Free, no sign-up, no limit on recordings. Choose English mode, or Any language mode that recognizes 99 languages with automatic detection, though accuracy is lower outside English. Up to 2 minutes per recording, as many recordings as you like. A one-time model download (about 80 MB, or about 165 MB with GPU acceleration) is cached by the browser, then recognition works without a connection. Works best in current desktop Chrome or Edge. Audio and text are never stored on a server, and there is no account. The site records anonymous usage events only.
Yap is a voice dictation app for Mac and Windows. We wanted a free way to try speech to text in a browser without installing anything, so we built this page. Recognition runs on your own device, so the audio is never uploaded.