OpenAI and Google have introduced new tools and models to advance artificial intelligence translation. OpenAI launched ChatGPT Translate, a standalone tool designed to compete with established services. It supports over 50 languages and uses a layout similar to existing translation platforms.
Features of ChatGPT Translate
A primary feature of ChatGPT Translate is its integration with generative AI. After a translation is generated, users can select options to adjust the tone or target audience. These presets include making text more formal or simplifying it for children. Selecting an option sends the text to the main ChatGPT interface for further refinement. Currently, the tool focuses on plain text for desktop users and voice input for mobile browsers. It lacks support for documents or website translations.
Google introduces TranslateGemma
Google announced TranslateGemma, a suite of open translation models built on its Gemma 3 architecture. These models are available in three different sizes to run on devices ranging from mobile phones to cloud servers. Google states that these models achieve high efficiency through a specialized training process involving both supervised fine-tuning and reinforcement learning.
TranslateGemma supports 55 core languages. Google claims the 12B parameter model outperforms some larger baseline models in accuracy. According to the company, TranslateGemma retains strong multimodal capabilities. This allows the models to translate text found within images more effectively. While OpenAI offers a consumer-facing tool, Google positions TranslateGemma as a foundation for researchers and developers to build upon.
Sources: Android Authority, Google Blog
