Google attracts Indian developers with new AI models and tools
Google is setting its sights on India's burgeoning artificial intelligence (AI) developer community, unveiling a suite of tools, programs, and partnerships at the I/O Connect 2024 event in Bengaluru today. The tech behemoth aims to aid developers in crafting AI-driven products and solutions for both domestic and international markets. "India is at the cornerstone of our global AI mission," said Seshu Ajjarapu, Senior Director at Google DeepMind.
India's prominent role in Google's AI mission
Ajjarapu emphasized that India is a global leader in adopting Google's Gemini family of AI models, utilized by over 1.5 million developers worldwide. He also pointed out that India boasts one of the largest user bases for Google's online developer platform, Google AI Studio. Manish Gupta, Director at Google DeepMind, identified multimodal, multilingual, and mobile as key areas of focus for the company's AI initiatives.
Google announces release of Gemma 2 for Indian developers
During the event in Bengaluru, Google announced the release of Gemma 2 to all developers in India. This next-generation open-source AI model features a new architecture designed for improved performance and efficiency. Available in both 9 billion and 27 billion parameter sizes, Gemma's tokenizer is particularly effective for building multilingual solutions that cater to India's diverse languages.
Google's Gemma powers multilingual solutions for Indian languages
Gemma's effectiveness was demonstrated by Navarasa, a multilingual variant for Indian languages developed by Telugu LLM Labs. Built on Gemma, Navarasa presently supports 15 Indian languages. Additionally, Google announced the availability of a 2 million token context window on its flagship Gemini 1.5 Pro model to all developers in India, enhancing data processing capabilities.
Google introduces IndicGenBench and CALM for Indian languages
Google is launching IndicGenBench, a multilingual benchmark suite specifically designed for Indian languages. Developed by Google DeepMind's India unit, the suite can evaluate language generation capabilities of large language models across various user-facing tasks in 29 Indian languages. The tech giant is also open-sourcing a new technology called CALM (Composition of Language Models) that allows developers to combine their specialized language models with Gemma models.
Google DeepMind India develops Matformer framework
The Google DeepMind India team has played a significant role in developing a new framework called Matformer, which is designed to enhance on-device AI capabilities. This framework will be included in the second version of Gemini Nano, which is expected to be released soon. Google's collaboration with MeitY 'Startup Hub' aims to train 10,000 startups in AI, further expanding access to its powerful AI models.
Google unveils Wallet APIs and offers discounted pricing
Google has also unveiled Google Wallet APIs to simplify the integration of loyalty programs, tickets, and gift cards. For developers using the Google Maps platform, the company is introducing India-specific pricing with up to 70% lower costs on most APIs. Additionally, in collaboration with the Open Network for Digital Commerce (ONDC), Google is providing developers building for ONDC up to 90% off on select Google Maps APIs.
