Google's WebSpeech API looks pretty good: https://www.google.com/intl/en/chrome/demos/speech.html https://www.google.com/intl/en/chrome/demos/speech.html CMU S

Google's WebSpeech API looks pretty good: https://www.google.com/intl/en/chrome/demos/speech.html https://www.google.com/intl/en/chrome/demos/speech.html CMU Sphinx and Julius both also have options for continuous speech recognition, but it's quite a feat to train your own [good] language model.