Skip to content

Changelog

This page documents changes in the Streaming Speech-to-Text API only. View all changes.

2022-10-27

Changed

  • Streams are now billed based on the maximum of stream duration and audio duration. Refer to the Billing section for more details.

2022-09-29

Added


2022-09-22

Added

  • Added support for Streaming Speech-to-Text API (English only) in the European Union deployment.

2022-08-04

Added

  • Added a new priority option to the Streaming Speech-to-Text API. Possible values are speed and accuracy. Only available for English and Spanish languages with machine_v2 transcriber.

2022-07-25

Changed

  • Streaming Speech-to-Text API v2 moved from Open Beta to General Availability

2022-05-26

Added

  • Added a new skip_postprocessing option to the Streaming Speech-to-Text API. This option allows you to skip the post-processing steps (inverse text normalization or ITN, casing and punctuation) of a transcription job. Only available for English and Spanish languages.

2022-05-26

Added

  • Added machine_v2 transcriber. Routes to our new improved v2 model.
  • Added enable_speaker_switch option. Only available for v2 streams.

2022-05-20

Changed

  • The 8 languages for the Streaming Speech-to-Text API are out of Open Beta and in General Availability: French, German, Italian, Japanese, Korean, Mandarin, Portuguese, and Spanish.

2022-04-25

Added

  • Added 8 new languages to the Streaming Speech-to-Text API in Open Beta: French, German, Italian, Japanese, Korean, Mandarin, Portuguese, and Spanish.

2022-04-20

Added

  • RTMP audio streaming documentation

2021-05-06

Added

  • Enable offsetting hypotheses timestamps by providing start_ts to streaming jobs

2020-08-27

Added

  • (Open Beta) detailed_partials option to show timestamps and confidence scores in partial hypotheses

2020-06-24

Added

  • (Closed Beta) Option to show timestamps and confidence scores in partial hypotheses. Email support@rev.ai for access.
  • (Closed Beta) Stream to Rev AI with RTMP. Email support@rev.ai for access.