AI service
Amazon Transcribe
Converts speech audio to text transcripts.
Key points
- Converts audio or video speech into text transcripts.
- Supports batch and streaming transcription use cases.
- Can identify speakers, add timestamps, and use custom vocabularies in supported workflows.
- Includes features for redaction and call analytics in relevant variants.
- Produces text that can be analyzed by services such as Comprehend or searched later.
When to use it
- Choose Transcribe for meeting notes, call transcripts, captions, or voice analytics.
- Use streaming transcription for near real-time captions or voice applications.
- Use custom vocabulary when domain terms, product names, or acronyms must be recognized.
Exam tips
- Transcribe is speech-to-text; Polly is text-to-speech.
- Lex is for conversational bots, while Transcribe is focused on transcription.
- Comprehend can analyze the resulting transcript for sentiment, entities, or PII.
- For multilingual text output, combine Transcribe with Translate when appropriate.