Identify speech segments in audio recordings
A dual reasoning model that runs locally in your browser.