AI Glossary

Speech-to-text

Also called speech recognition, ASR, transcription

What does “Speech-to-text” mean?

Speech-to-text: Turning spoken audio into written text. Systems are trained on recordings paired with transcripts, and current models handle accents, background noise and many languages far better than older ones; some also identify who is speaking or add timestamps. Errors still occur with rare names, jargon and overlapping voices.

Speech-to-text visual guide: Turning spoken audio into written text. Systems are trained on recordings paired with transcripts, and current models handle accents, background noise and many languages far better than older ones; some also identify who is speaking or add timestamps. Errors still occur with rare names, jargon and overlapping voices.
Visual guide Open full size (new tab)

What is an example of “Speech-to-text”?

OpenAI's open-weight Whisper model is widely used for transcribing meetings, generating subtitles and powering voice input to chatbots.

Keep learning

See also

Text-to-speechTermNatural language processingTermMultimodal AITermLip syncTerm

Same topic

Style transferTermSeedTermTemporal consistencyTermSampling stepsTermText-to-imageTerm