Tuesday, 30 September, 2025
Nasscom Plans Local Benchmarks for Indic AI Models to Address Global Testing Gaps
Nasscom is initiating the development of localized benchmarks to evaluate AI models tailored for Indian languages. The current global benchmarks, such as MMLU and BIG-bench, predominantly focus on English and European languages, often overlooking the complexities of code-mixed and multilingual contexts prevalent in India. Nasscom's initiative aims to create an Indian-centric framework that includes multilingual, multimodal, and culturally relevant use cases, ensuring fair evaluation.
Read full story at Economic Times