From Indic languages to JEE mock tests: How Google is scaling AI offerings for Indian users

Google on Friday shared the latest developments across its artificial intelligence products at the AI Impact Summit, highlighting how earlier launches are being expanded globally and in India to make its tools more helpful in everyday life.

In a post on X, the tech giant said it is focused on breaking language barriers and improving access to knowledge, outlining recent enhancements in real-time translation, visual search, student learning and online safety.

Google said its live speech-to-speech translation model now supports real-time conversations in more than 70 languages, including 10 Indic languages such as Hindi, Bengali, Tamil and Telugu.

The company has also rolled out AI-powered speaking and listening practice features designed to help users learn new languages, tailored to individual proficiency levels.

It added that in the coming weeks it will enhance the model behind Search Live, its voice- and camera-based search tool. The improved model will enable more users to ask questions about what they see in their own language, including Hindi, Indonesian, Japanese, Korean and Brazilian Portuguese.

“We're also testing ways to make it easier to get helpful information from visual search even when the images aren't clear,” Google said.

Google said it has recently introduced new self-study features and JEE Main practice tests for Indian students on its Gemini app, with similar capabilities coming soon to AI Mode in Search.

The tech giant has grounded practice tests in “rigorously vetted content” from ed-tech firms like PhysicsWallah and Careers36.

“This helps ensure that you’re not just practicing — you’re preparing with material that more closely resembles what you’ll see on test day,” Google had said in its workspace updates.

Students can initiate mock exams through simple prompts, receive scores and skill-wise feedback, and access step-by-step explanations and personalised study plans.

Google also said its SynthID verification feature in the Gemini app has been used over 20 million times across languages to help identify AI-generated images, video and audio.

Google also shared its annual Responsible AI progress report, detailing efforts to strengthen safety and security across its AI systems.