News Brief

AI App ‘Adi Vaani’ Launched To Preserve And Translate Tribal Languages, Starting With Bhili, Gondi, Santali, Mundari

Arun DhitalSep 02, 2025, 12:38 PM | Updated 12:38 PM IST
Adi Vaani App Launch (X/@TribalAffairsIn)

Adi Vaani App Launch (X/@TribalAffairsIn)


The Ministry of Tribal Affairs on Monday (1 September) launched the beta version of ‘Adi Vaani’, an AI-powered translation tool designed to digitise, translate, and preserve India’s tribal languages, the Indian Express reported.

Developed by a consortium led by IIT-Delhi with BITS Pilani, IIIT Naya Raipur, and tribal research institutes in Jharkhand, Odisha, Madhya Pradesh, Chhattisgarh, and Meghalaya, the app marks a major step in safeguarding endangered linguistic heritage.

The app allows translation between Hindi, English, and tribal languages, and will initially support Bhili, Gondi, Santali, and Mundari, four of the most widely spoken tribal tongues.

Bhili alone is spoken by over 1 crore people across central and western India. Subsequent updates will include Kui of Odisha and Garo, spoken in the Northeast.

India has 461 tribal languages and 71 distinct mother tongues, according to the 2011 Census, but many are at risk: 82 are classified as vulnerable and 42 critically endangered.

Adi Vaani aims to bridge this gap by creating a digital corpus for these languages by building a model with extensive community participation.


The app also integrates tribal language primers developed by NCERT and the Central Institute of Indian Languages, Mysuru, covering 117 languages, including 89 tribal ones.

These primers are expected to aid foundational learning in mother tongues.

Ministry Secretary Vibhu Nayar said the initiative will not only preserve linguistic traditions but also improve governance and education delivery.

Features like an optical character recognition (OCR) tool will enable better access to schemes and study materials for tribal students. The app will be available on the Google Play Store from Friday.

Join our WhatsApp channel - no spam, only sharp analysis