Quick teaser of what I’ve been working on over the last few weeks: a streaming medical speech-to-text model that runs fully on-device. This demo is running locally on a MacBook through MLX. Still doing more evals, but planning to release the open weights next week.
