Adi Vaani: AI Platform Supports Nine Tribal Languages and Community-Led Digital Preservation

The news

The Ministry of Tribal Affairs’ artificial intelligence-powered platform Adi Vaani supported text translation in nine tribal languages and recorded over 35,000 Android downloads as of 15 October 2026, the ministry said. Launched on 1 September 2025, the platform was developed by a national consortium led by the Indian Institute of Technology Delhi, along with BITS Pilani, IIIT Hyderabad, IIIT Nava Raipur and five State Tribal Research Institutes. The Working Group on Tribal Languages, which first met on 15 October 2026, discussed positioning Adi Vaani as the National Tribal Language Repository, though this remains a proposal.

State Tribal Research Institutes have contributed approximately 13.9 lakh text sentences across ten languages and around 300 hours of speech data to train the platform’s models, the ministry said. Adi Vaani works with 13 tribal languages overall, including Santali, Bhili, Mundari, Gondi, Garo, Kui, Kokborok, Betta Kuruba, Katkari, Koya, Thadou Kuki, Tangkhul and Chodri. The platform also provides speech translation, Optical Character Recognition, video subtitling and access to more than 300 language primers. It has around 6,500 monthly users, according to the ministry.

When was the Ministry of Tribal Affairs established?

The Ministry of Tribal Affairs was carved out of the Ministry of Social Justice and Empowerment as a separate ministry in 1999. Tribal Research Institutes were set up under the ministry to act as state-level knowledge and research centres for tribal development. The Constitution of India, under Article 342, provides for the specification of Scheduled Tribes, and Article 46 directs the State to promote the educational and economic interests of Scheduled Tribes. The National Education Policy 2020, which recommends home-language instruction, was approved by the Union Cabinet in July 2020.

How does the platform collect and validate data?

Native speakers contribute words, sentences and voice recordings through State Tribal Research Institutes and AdiKosh, the platform’s word bank. Language experts review the contributions before they are used to train AI models. The project aims to collect around 1,00,000 sentence pairs and 100 hours of speech recordings per language. Each language passes through five phases, from data collection to integration on the platform, and the models are retrained as the community contributes more data.

What are the potential applications beyond translation?

The National Education Policy 2020 recommends home or local language instruction wherever possible until at least Grade 5, preferably Grade 8 and beyond. Adi Vaani resources have potential use across more than 525 Eklavya Model Residential Schools serving over 1.7 lakh students, the ministry said. Translated sickle cell awareness materials and accessible market information offer additional health and livelihood applications. The platform can also support access to public services by reducing language barriers for tribal communities.

What is the next step for the platform?

The Working Group on Tribal Languages considered a national language baseline, digital archives, multilingual educational material and district-level community resource groups, the ministry said. A proposed two-year action plan includes quarterly reviews and stronger collaboration among research institutes, academic institutions and language experts. The Ministry of Tribal Affairs will circulate agenda items and prepare a roadmap; these measures remain proposals.