Amazon Transcribe features

To help you decide which Amazon Transcribe solution best fits your use case, the following table offers a feature comparison.

Note that 'batch' and 'post-call' refer to transcribing a file that is located in an Amazon S3 bucket and 'streaming' and 'real-time' refer to transcribing media in real time.

Feature	Amazon Transcribe	Amazon Transcribe Medical1	Amazon Transcribe Call Analytics
Configuration options
Alternative transcriptions	batch, streaming	batch, streaming	no
Channel identification	batch, streaming	batch, streaming	post-call, real-time
Job queueing	batch	no	post-call
Language identification	batch, streaming	no	post-call
Multi-language identification	batch, streaming	no	no
Speaker diarization	batch, streaming	batch, streaming	post-call
Transcribing digits2	batch, streaming	batch, streaming	post-call, real-time
Conversation analytics
Call characteristics	no	no	post-call
Call summarization2	no	no	post-call
Custom categorization	no	no	post-call
Real-time category events	no	no	real-time
Real-time issue detection2	no	no	real-time
Real-time speaker sentiment	no	no	real-time
Speaker sentiment	no	no	post-call
Language customization
Custom language models2	batch, streaming	no	post-call, real-time
Custom vocabularies	batch, streaming	batch, streaming	post-call, real-time
Resource organization
Tagging	batch	batch	post-call
Sensitive data
Identifying personal health information2	no	batch, streaming	no
Identifying personally identifiable information2	streaming	no	real-time
Redacting audio2	no	no	post-call, real-time
Redacting transcripts2	batch, streaming	no	post-call, real-time
Vocabulary filtering	batch, streaming	no	post-call, real-time
Video
Subtitles	batch	no	no

¹ Amazon Transcribe Medical is only available in US English.

² This feature is not available for all languages; review the Supported languages and language-specific features table for more details.