Home › Learn › What Is Real-Time Transcription? — How It Works
Real-time transcription converts speech to text as it is spoken, with minimal delay. Unlike batch transcription (which processes a recording after it is finished), real-time transcription shows words appearing on screen within milliseconds of being spoken.
How real-time transcription works
The audio is streamed to the AI model as it is captured
The model processes audio in small chunks (typically 100-300ms)
Words appear on screen within 300-500ms of being spoken
The transcript is refined as more context becomes available
Speaker labels are applied in real time
Real-time transcription in Talk2Memo
Talk2Memo transcribes in real time — words appear on screen within 300ms of being spoken. You can see a live transcript during any meeting, interview, or lecture. After recording, the full transcript is refined and an AI summary is generated.
Try real-time transcription for free
Words appear within 300ms of being spoken. Free plan: 200 minutes per month.
Start Free — No Card Required