Automatic speech recognition (ASR)
The technology that converts spoken audio into text automatically. Modern ASR is powered by deep-learning models such as Whisper.
Automatic speech recognition (ASR) is the underlying technology behind transcription: software that turns speech into text without a human typist. Today’s ASR is driven by deep-learning models — Whisper-class models are the common baseline — that reach low single-digit word error rates on clean audio.
ASR produces the raw words; adding diarization, punctuation and structure is what turns its output into something useful.