The BRICS TimesDeveloped using AI and contributions from educators fluent in tribal dialects, the app enables two-way translation between Hindi, English, and tribal languages, while working to digitize and safeguard them.
The Ministry of Tribal Affairs on Monday introduced the Beta version of Adi Vaani, an AI-driven translation platform designed to support large language models for India’s tribal languages.
Developed by a consortium led by IIT Delhi, alongside BITS Pilani, IIIT Naya Raipur, and several tribal research institutes across Jharkhand, Odisha, Madhya Pradesh, Chhattisgarh, and Meghalaya, the app is both a translation tool and a language preservation initiative.
India has a Scheduled Tribe population of over 10 crore, speaking 461 tribal languages, including 71 distinct mother tongues. Many of these are endangered—with 82 considered vulnerable and 42 critically endangered. To begin with, Adi Vaani will offer translations for Bhili, Gondi, Santali, and Mundari, before expanding to include Kui and Garo. The app will be available on Google Play Store from Friday.
Creating a linguistic database was a major hurdle for developers. More than 250 tribal speakers, educators, and community leaders contributed by compiling dictionaries, translating NCERT textbooks, and recording folklore and oral traditions to build the corpus.
The platform also incorporates primers created under the Ministry of Education’s initiative, covering 117 languages (including 89 tribal ones), with Santali, Warli, and Pawri already integrated.
According to Vibhu Nayar, Secretary of the Tribal Affairs Ministry, the translation system will enhance the implementation of government schemes at the grassroots level and make educational resources more accessible, especially with its OCR (optical character recognition) feature.
With the ongoing Adi Karmayogi Abhiyan, aimed at building local tribal leadership, the ministry expects Adi Vaani to gain traction among tribal communities.
