We're excited to announce a new feature in the Dictation mode: Semantic mode ("Send sentences automatically when complete"). This new option uses advanced AI to intelligently determine the end of phrases, providing a more natural and accurate transcription experience.

• What is Semantic Segmentation? Unlike the traditional time-based segmentation which cuts off a phrase after a set period of silence, semantic segmentation uses AI to understand the meaning and flow of a sentence. This feature is especially beneficial for fast speakers, as it ensures that phrases are not prematurely cut off.

• How to Use It: You can now choose between "Semantic" and "Time" segmentation types in the speech-to-text settings. Simply navigate to the settings and select your preferred method.

• Automatic Fallback: If you select "Semantic" but the language you are using is not yet supported, the app will automatically revert to the "Time" based segmentation to ensure a seamless experience.

• Supported Languages: Semantic Segmentation is currently available for the following languages:

  • English (en-US, en-GB)

  • Chinese (zh-CN, zh-HK)

  • Japanese (ja-JP)

  • Korean (ko-KR)

  • German (de-DE)

  • French (fr-FR)

  • Italian (it-IT)

  • Spanish (es-ES, es-MX)

  • Hindi (hi-IN)

  • Portuguese (pt-BR, pt-PT)

  • Turkish (tr-TR)

  • Russian (ru-RU)

  • Thai (th-TH)

  • Indonesian (id-ID)