Bodhan AI launches 4 open models for multilingual education

CHENNAI: Bodhan AI, the Centre of Excellence in AI for Education incubated at IIT-Madras, launched four AI models for Indian languages, making them available as Digital Public Goods to help institutions, startups, and researchers build applications.

Developed in partnership with AI4Bharat, the models cover speech recognition, speech generation, machine translation and optical character recognition, with support ranging from 22 to 27 Indian languages. They are being offered through open-weight releases and APIs on sovereign digital infrastructure, enabling users to adapt the technology without developing models from scratch.

The models are part of the Bharat EduAI Stack, envisioned as sovereign digital public infrastructure for education.

"India's AI journey cannot be built on technology alone. It must be built on technology that understands India," said IIT Madras Director V Kamakoti.

Bodhan AI Principal Investigator Mitesh Khapra said the initiative aimed to build with the ecosystem, not compete with it, creating a common AI layer and reducing duplication.

The launch also includes Student TutorBot and Teacher Assistant Bot.

TutorBot is designed for students in Classes 6-12 and supports text and voice in 22 Indian languages, using textbooks for explanations, practice and assessment. The Teacher Assistant Bot can help educators prepare lesson plans, worksheets, quizzes and assessments, while leaving teachers in control of reviewing or discarding AI-generated outputs.

The models were trained and optimised using NVIDIA Nemotron open models and are intended to support AI applications across India's linguistic and educational landscape.