(MENAFN- KNN India) Mumbai, Nov 26 (KNN) Union Minister of State for Science & Technology Dr. Jitendra Singh praised BharatGen, calling it India's first sovereign multilingual and multimodal Large Language Model (LLM), during his visit to IIT Bombay.

BharatGen is being built as a national AI infrastructure supporting over 22 Indian languages across text, speech, and document-vision formats. The project aligns with India's push for indigenous deep-tech and culturally rooted AI systems.

The project is funded with Rs 235 crore under the National Mission on Interdisciplinary Cyber-Physical Systems (NM-ICPS) programme through IIT Bombay's Technology Innovation Hub.

It also includes Bharat Data Sagar, a sovereign dataset repository aimed at strengthening India's digital autonomy.

The minister reviewed several released models, such as Param-1 (a 2.9B-parameter text model), Shrutam (speech recognition), Sooktam (text-to-speech in nine languages), and Patram (India's first document-vision model).

Demonstrations included Krishi Sathi for voice-based farm advisories and e-VikrAI, which generates product descriptions from a single image. Docbodh, a Patram-powered platform that simplifies complex documents, was also showcased.

BharatGen is collaborating with IBM, Zoho, NASSCOM, various ministries including the Department of Drinking Water and Sanitation, and state governments like Maharashtra.

Recently, the project secured an additional Rs 1,058 crore under the India AI Mission, bringing total government support to Rs 1,293 crore.

Calling BharatGen a major step toward technological self-reliance, Dr. Singh said the initiative reflects the vision of inclusive, India-centric AI, and urged the team to continue building globally competitive yet widely accessible models.
(KNN Bureau)

MENAFN26112025000155011030ID1110399329