Comment by rason-yang

I build low-latency voice gateways . Does your local API expose interim transcripts and timestamps while audio is arriving , or only the final text after a dictation ends ?