CHENNAI: Bodhan AI, a Centre of Excellence in Artificial Intelligence for Education incubated at IIT Madras, has launched four AI models designed to support Indian languages and build technology for multilingual education.
Developed in partnership with AI4Bharat, the models cover speech recognition, optical character recognition, machine translation and text-to-speech. Depending on the application, the systems can work across between 22 and 27 Indian languages.
The models are being released as open-weight systems and through hosted application programming interfaces, allowing educational institutions, researchers and technology developers to build applications without having to develop similar AI capabilities from the beginning.
The initiative is part of efforts to create a common technology layer for education applications that can work across India’s diverse linguistic environment.
Bodhan AI is also developing tools that use these models for students and teachers. A multilingual student tutor is designed to help learners with academic questions, while a teacher assistant is intended to support activities such as lesson planning, assignments and evaluation.
The AI systems are being developed with education-focused applications in mind, including support for Indian languages, regional accents and classroom-related content.
The initiative could help make digital learning tools more accessible to students who are more comfortable using Indian languages rather than English.
Multilingual AI has become an important area of technology development in India as digital education expands across different regions and language communities.
By making the underlying models available for wider use, the initiative also aims to encourage researchers, educational technology companies and other developers to create applications suited to Indian requirements.
The launch marks another step towards developing AI systems specifically for India’s linguistic and educational needs.
Comment here