iSpeech's open source javascript SDK for speech recognition (ASR) API, enables you to easily create Web applications using iSpeech freeform, command or custom statistical language models. The speech recognition API powering this speech recognition SDK supports nearly 30 languages and accents. The acoustic models are based on huge amounts of low and high quality hand labeled audio data (millions of utterances). iSpeech is a viable alternative to Google ASR (Web Speech API), which only includes Voice Search and Freeform Models, or others that do not offer a Javascript SDK to our knowledge. -Currently this SDK has been tested and works on Chrome and Firefox, but not Internet Explorer (IE) or Safari - (has not been tested on Opera browser) *you may need to contact iSpeech to create custom models **if you do not have an iSpeech account and need to test, you may request credits *** iOS, Android, Java, .NET, Python and other SDKs are also available