Google on Wednesday announced a host of developer-centric updates along with news around its Gemini and Gemma AI models at its I/O Connect event in Bengaluru. Google DeepMind’s India team also revealed multiple initiatives including Project Vaani, a collaboration with the Indian Institute of Science which will focus on capturing over 14,000 hours of speech data across 58 languages, collected from 80,000 speakers in 80 districts for developers.
The team has developed IndicGenBench under the project, a benchmark developers can use to evaluate the capabilities of large language models based on indic languages. Additionally, they have also open-sourced a mix of specialised LLMs with Google’s lightweight Gemma AI models, called the Composition of Language Models framework or CALM.
India’s developer relations lead Karthik Padmanabhan announced the company will be partnering with the India’s Ministry of Electronics and Information Technology (MeitY) to support 10,000 Indian generative AI startups. They will be offered Google Cloud credits, an updated startup program curriculum to equip founders with the necessary skills required to develop AI. Additionally, Google will also host an annual nationwide GenAI Hackathon and AI Startup Bootcamp.
Google also confirmed that developers in India will finally be able to access the expanded context window with up to 2 million parameters on the latest Gemini 1.5 Pro model and Gemma 2.
(For top technology news of the day, subscribe to our tech newsletter Today’s Cache)
For developers on the Google Maps Platform, the company has introduced a specific pricing with up to 70 percent lower costs on most APIs.
Google has also partnered with the Open Network for Digital Commerce (ONDC) to offer developers building on the platform up to 90 percent on select Google Maps Platform APIs.
A limited availability tool called the Agricultural Landscape Understanding or ALU Research API will be launched soon with the aim to make agricultural practices more data-driven and efficient.
“India is at the forefront of the AI revolution, as you can see from the innovation that Indian companies are pioneering. From consumer experiences to agriculture to social enterprises, AI has the power to address some of the biggest challenges of our time across many sectors and industries,” said Seshu Ajjarapu, Senior Director, Google DeepMind.
Meanwhile, Project Oscar, an AI agent, has been designed to help open-source developers keep track of their latest projects. The AI agent will be initially made for Go, which has more than 2,000 contributors currently and over 93,000 commits i.e., changes made to files. Padmanabhan added that Project Oscar can also be adapted to support other open-source projects.