On Wednesday, Google held its I/O Connect India 2025 event in Bengaluru, where it introduced a suite of AI-driven tools and initiatives designed specifically for the Indian developer community. The event featured presentations from eight domestic startups that have utilized Google’s advanced AI models to enhance their applications. Among them, Bengaluru-based Sarvam unveiled its new translation tool, built on Google’s Gemma 3 model, while other startups, including CoRover, Dashverse, Entri, Glance, Invideo, Nykaa, and Toonsutra, showcased innovations powered by Google’s Gemini technology.
Sarvam Leverages Gemma 3 for High-Volume Translations
During the event, eight Indian startups, namely CoRover, Dashverse, Entri, Glance, Invideo, Nykaa, Sarvam, and Toonsutra, presented their AI-developed solutions using Google’s frameworks.
Sarvam has launched Sarvam-Translate, an open-source translation model built with Gemma 3. This innovative tool is designed to efficiently deliver translations for long-form content across all 22 Indian languages. Google reported that the hosted translation API manages over 100,000 translation requests each week and also supports Sarvam’s multilingual conversational AI platform, Samvaad.
Another notable contributor, CoRover, utilized Google’s Gemini model to develop customizable, multilingual chatbots for businesses. Google indicated that CoRover’s BharatGPT AI model has reached over 1 billion users and partnered with more than 25,000 enterprises, handling upwards of 20 billion interactions, all made possible through the Google Cloud AI Stack.
Glance announced its use of both Gemini and Imagen technologies via Vertex AI to enhance user personalization. The platform enables users to upload selfies and generate various stylistic options based on their personal preferences.
In the education sector, Entri has integrated Gemini into its AI-powered Teacher Assistant and Interview Coach tools, offering real-time assistance in multiple Indian languages. According to Google, over 53% of users have accessed the Teacher Assistant, and 91% have utilized the Interview Coach feature.
Moreover, Invideo is revolutionizing video production with the application of Gemini models like Imagen 4 and Veo 3, allowing users to convert text prompts into complete videos while Google’s AI stack manages the technical aspects.
Retail giant Nykaa employs Gemini through Chrome’s multimodal API to facilitate visual product searches on its fashion platform, enabling users to take pictures and find corresponding or suggested items from their catalogue.
Other startups introducing Google’s AI models include Dashverse and Toonsutra. Dashverse has integrated Gemini, Veo 3, and Lyria 2 on Vertex AI to enhance its platforms, Dashtoon Studio and Frameo, which transform text prompts into comics and cinematic videos. Their applications, Dashtoon and Dashreels, reportedly cater to over 2 million users.
Toonsutra is also harnessing the power of Gemini 2.5 Pro and Lyria 2 to create Indian-language webcomics enriched with music and character voices, utilizing Veo 3’s image-to-video capability to animate comic visuals.
Additionally, Google announced the launch of the second cohort of its Google for Startups Accelerator: Apps programme, featuring 20 AI-focused startups from India. Over the last ten years, Google’s accelerator programmes have reportedly supported more than 230 startups across the country.
