Summary
Phonon-2 is a compact open English speech-recognition model for local transcription. Its encoder stores weights in five learned levels at approximately 2.1 bits, compressing a Parakeet-derived model into a 164 MB download.
Phonon-2 is a compact open English speech-recognition model for local transcription. Its encoder stores weights in five learned levels at approximately 2.1 bits, compressing a Parakeet-derived model into a 164 MB download.
Developers can transcribe recordings across desktop and GPU environments using the same model file and attribution-based license. Core capabilities include English speech recognition, 164 MB download, Learned low-bit encoder weights, and Parakeet-derived model.
Public code or model weights are linked from the project, with setup instructions for running or adapting the system. Additional characteristics include Mac compatibility, Linux and Windows support, NVIDIA GPU support, and CC-BY-4.0 weights.
Get more likes & reach the top of search results by adding this button on your site!


Get top updates in AI to your inbox every weekend. It's free!