.png "What Are Ebike ‘Classes’ and What Do They Mean?")
Google hosted its I/O Connect India 2025 event in Bengaluru on Wednesday, unveiling new AI-powered tools and programmes tailored for the Indian developer ecosystem. Eight domestic startups showcased applications made using Google’s new AI models during the event. Bengaluru-based AI startup Sarvam has leveraged Google’s Gemma 3 model for its latest translation tool. Similarly, startups including CoRover, Dashverse, Entri, Glance, Invideo, Nykaa, and Toonsutra have used Google’s Gemini to power their new features.
Sarvam Uses Gemma 3 to Process 1 Lakh Translation Requests Every Week
At Google I/O Connect India 2025, eight Indian startups showcased innovations built using Google’s AI models: CoRover, Dashverse, Entri, Glance, Invideo, Nykaa Sarvam, Toonsutra.
Bengaluru-based AI startup Sarvam has used Gemma 3 to develop Sarvam-Translate, its newly launched open-source translation model. This model is designed to deliver translations for long-form content across all 22 Indian languages. Offered as a hosted translation API, Google claims that it processes over 1,00,000 translation requests each week. The API also powers Samvaad, Sarvam’s multilingual conversational AI platform.
Another AI firm, CoRover, leveraged Google’s Gemini model to create customisable, multilingual customer chatbots for businesses. According to Google, CoRover’s BharatGPT AI model has served more than 1 billion users, over 25,000 enterprises and developers, and handled over 20 billion interactions. It used Gemini through the Google Cloud AI Stack.
Glance says it used Gemini and Imagen through Vertex AI to create a personalised experience for users. It allows users to upload selfies through smartphones to generate different looks styled on their own image based on their preferences.
Edtech platform Entri has integrated Gemini into its AI-powered Teacher Assistant and Interview Coach tools. This provides users with real-time support in Indian languages. Google states that over 53 percent of users accessed the Teacher Assistant, while 91 percent have tried the Interview Coach.
At Google I/O Connect India 2025, Google also revealed that Invideo is transforming video creation by using Gemini models like Imagen 4 and Veo 3. With this, the company lets anyone turn text prompts into full-length videos. Using Google’s AI stack, Invideo handles the technical side.
Retail firm Nykaa uses Gemini via Chrome’s multimodal API to enable visual product search on its fashion site. This lets users snap a photo to find matching and suggested items from the catalogue.
Other Indian startups joining Google’s AI models are Dashverse and Toonsutra. Dashverse used Gemini, Veo 3, and Lyria 2 on Vertex AI to power its AI platforms, Dashtoon Studio and Frameo. These AI platforms can transform text prompts into comics and cinematic videos. These tools drive its apps, Dashtoon and Dashreels, which are claimed to be serving over 2 million users.
Webcomics application Toonsutra is leveraging Gemini 2.5 Pro and Lyria 2 for creating Indian-language webcomics with music and character voices. It uses Veo 3’s image-to-video feature to bring comic visuals.
Additionally, Google announced the second cohort of its Google for Startups Accelerator: Apps programme, featuring 20 AI-driven Indian startups. The company claimed that over the past decade, Google’s accelerator programmes have supported more than 230 startups in India.
