The newly released IndicGenBench explores the generative capabilities of LLMs across 29 Indic languages, marking a significant step towards linguistic inclusivity in AI research. Key aspects include:

• Diverse Tasks: Involves machine translation, summarization, and question answering.
• Cross-Lingual Evaluation: Provides a unique cross-lingual benchmark improving representation for under-represented languages.

This benchmark is crucial for driving future research in multilingual AI, potentially enhancing models’ effectiveness across a broader range of linguistic and cultural contexts.