Maharashtra
BharatGen
- 26 Nov 2025
- 3 min read
Why in News?
BharatGen, India’s first sovereign, multilingual and multimodal AI-driven Large Language Model (LLM), was highlighted by Union Minister of Science and Technology during his visit to IIT Bombay.
Key Points
- About BharatGen:
- BharatGen is India’s first government-supported sovereign AI stack, developed to support over 22 Indian languages across text, speech, and document-vision modalities.
- It has been developed by a consortium led by IIT Bombay, in collaboration with IIT Madras, IIT Kanpur, IIT Hyderabad, IIT Mandi, IIT Kharagpur, IIIT Hyderabad, IIIT Delhi, and IIM Indore.
- The project is funded with ₹235 crore under National Mission on Interdisciplinary Cyber Physical Systems (NM-ICPS), along with an additional ₹1,058 crore from MeitY under the India AI Mission, bringing the total public investment to ₹1,293 crore.
- It is designed to advance India’s digital sovereignty, ensuring that AI development is rooted in Indian languages, cultural contexts, and national priorities.
|
Core AI Models Under BharatGen |
||
|
Model Name |
Modality |
Key Specifications |
|
Param-1 |
Text (LLM) |
Foundational text model trained on 7.5 trillion tokens, with one-third of the dataset comprising Indian linguistic data. |
|
Shrutam |
Speech (ASR) |
Automatic Speech Recognition (ASR) model designed to handle complex Indian linguistic diversity, including multiple dialects. |
|
Sooktam |
Speech (TTS) |
Text-to-Speech (TTS) model providing speech synthesis in nine Indic languages. |
|
Patram |
Document Vision |
India’s first document-vision model trained to interpret complex India-specific document formats such as identity records and legal documents. |
- About Bharat Data Sagar:
- Bharat Data Sagar is one of India’s largest sovereign data initiatives, aimed at ensuring complete national ownership of AI-relevant datasets.
- It focuses on India-centric data curation, covering diverse languages, dialects, cultural nuances, and region-specific contexts.
- It ensures data sovereignty, accuracy, and long-term national regulation over India’s AI ecosystem.
