Programmable Voice | Nov. 30, 2022

Updates to IBM Watch Speech to Text AddOn Language Models

The previous-generation language models that are currently available with IBM Watson Speech to Text AddOn will be deprecated and stop functioning as of March 3rd 2023. To avoid any issues with transcribing your call recordings use IBM Speech To Text AddOn, please update to use next-generation language models as soon as possible.

The following table has the mapping between previous generation language models, and the next generation language models to make it easier for you to make appropriate changes. We recommend that you make this change at your earliest convenience.

Previous Generation Model

Next Generation Model

en-US_NarrowbandModel

en-US_Telephony

en-GB_NarrowbandModel

en-GB_Telephony

es-ES_NarrowbandModel

es-ES_Telephony

pt-BR_NarrowbandModel

pt-BR_Telephony

zh-CN_NarrowbandModel

zh-CN_Telephony

While you can continue to use the previous generation models until March 3rd 2023, going forward you will only be able to select next generation models with any new installation of IBM Watson Speech to Text Addon


To learn more about the changes, please visit IBM Speech to Text documentation.

Voice GA