Google Translate prepares speed vs accuracy modes for translation, the AI model picker feature

Google is preparing a major AI-powered update for its Translate app, version 9.15.114, which introduces a new AI model picker feature. This will allow users, for the first time, to choose between translation modes optimized for speed or accuracy. The two options available will be:

  • Fast mode: Optimized for speed and efficiency, providing quicker translations using on-device processing.
  • Advanced mode: Powered by Google’s Gemini AI model, focusing on accuracy with more computationally intensive, cloud-based translations.

Currently, the advanced mode is limited to English-Spanish and English-French language pairs. The model picker will be accessible both on the main screen and on the translation results page.

This update includes reorganizing the UI for better accessibility, moving buttons for microphone, handwriting, and paste functions to a lower toolbar for easier one-handed use. The update aligns with broader AI trends of letting users balance speed and accuracy according to their needs, improving user control over translations from casual chats to professional documents.

This evolution in Google Translate builds on its Gemini AI capabilities and addresses user feedback for more reliable and context-aware translations, while also considering privacy by enabling some processing on-device. The rollout date for this feature to all users is not yet announced.

Comments

Leave a Reply

Your email address will not be published. Required fields are marked *