IIT Madras’ AI4Bharat Introduces New Benchmark Test for Indian Languages

AI4Bharat, a research lab at IIT Madras, has launched Indic LLM-Arena, an open-source benchmark to evaluate large language models (LLMs) on Indian languages, context, and safety. Developed with Google Cloud, it uses crowd-sourced voting from thousands of anonymous users to rank models on a “human-in-the-loop” leaderboard. Currently supporting text inputs and code-mixed language scenarios, plans include expanding to vision, audio, and AI agents. All data and code will be open-source. The platform aims to address the lack of local benchmarks, helping startups and researchers compare Indic LLM performance and enabling businesses to make informed AI adoption decisions.