Indian Languages · BHASHINI

The Government of India created the category of Classical Languages on 12 October 2004, declaring Tamil the first; Sanskrit followed in 2005, Telugu and Kannada in 2008, Malayalam in 2013 and Odia on 1 March 2014. A Linguistic Experts Committee under the Sahitya Akademi, constituted in November 2004, examines proposals. Maharashtra sought the status for Marathi in 2013, and during inter-ministerial consultations in 2017 the Ministry of Home Affairs advised that the criteria be revised and made stricter. The National Education Policy 2020, approved on 29 July 2020, gave Indian languages a larger place in schooling, with the mother tongue or regional language as the medium of instruction at least till Grade 5.

On 3 October 2024 the Union Cabinet approved Classical Language status for Marathi, Pali, Prakrit, Assamese and Bengali, taking the total from 6 to 11. The Linguistics Experts Committee had revised the criteria on 25 July 2024: high antiquity of early texts or recorded history over 1,500–2,000 years, a body of ancient literature considered a heritage by generations of speakers, and knowledge texts including prose, epigraphical and inscriptional evidence. The primary states involved are Maharashtra, Bihar, Uttar Pradesh, Madhya Pradesh, West Bengal and Assam. The Ministry of Culture said the decision would create employment in academic and research fields and aid the preservation, documentation and digitisation of ancient texts.

BHASHINI is run by the Digital India BHASHINI Division, under the Digital India Corporation of the Ministry of Electronics and Information Technology, through the National Hub for Language Technology. It offers speech and text AI services, including automatic language detection, speaker diarization and keyword spotting, which government websites and apps use to work in Indian languages. Its BhashaDaan crowdsourcing initiative lets citizens contribute and validate language datasets. AICTE's Anuvadini platform translates academic and technical content, including textbooks, into Indian languages, and Adi Vaani began with tribal languages such as Santali, Bhili, Mundari and Gondi.

BHASHINI's use grew through 2026: total inferences passed 6 billion by 12 March, and by 2 July it was processing more than 20 million inferences a day and had passed 8 billion. At the National Conference on e-Governance 2026 it launched a Rajasthan Language Model Training Hackathon to strengthen AI models for Rajasthan's regional languages. The Bharatiya Bhasha Pustak Scheme, proposed in the 2025-26 Budget, is meant to give school and higher education students digital books in Indian languages. The Gyan Bharatam Mission, approved with an outlay of ₹482.85 crore for 2024–31, uses AI tools to digitise and catalogue manuscripts.

6 → 11 classical languages (3 October 2024). 36 Indian text, 23 Indian voice and 35 international languages on BHASHINI (2 July 2026). 8 billion+ cumulative inferences and 20 million+ a day (2 July 2026). 800+ government websites powered (2 July 2026). 350+ optimised AI models (12 March 2026). 22 Scheduled and 99 Non-Scheduled languages (Census 2011). Sources: Ministry of Culture, Ministry of Education, MeitY (Digital India BHASHINI Division), Union Budget 2025-26, via PIB.