Overview
Bhashini — BHASHa Interface for India — is India’s first shared Digital Public Infrastructure (DPI) for language AI, launched in July 2022 under the National Language Translation Mission (NLTM). It uses cutting-edge Artificial Intelligence and Natural Language Processing (NLP) technologies to provide real-time translation, speech recognition, and text-to-speech services across all 22 Scheduled Languages of India and select tribal languages.
Bhashini is implemented by the Digital India BHASHINI Division (DIBD) under the Digital India Corporation (DIC) — a Section 8 Company of the Ministry of Electronics and Information Technology (MeitY). It aims to democratise access to digital content and governance services for India’s 1.4 billion citizens in their preferred language.
Key Statistics (as of early 2026)
| Parameter | Figure |
|---|---|
| Languages supported | 22 Scheduled Languages + tribal languages |
| AI capabilities | ASR, MT, TTS, Speech-to-Speech, OCR, Transliteration |
| Daily engagements (SabhaSaar) | 5 lakh+ |
| Total interactions (SabhaSaar since 2025 rollout) | 15.6 crore+ |
| Platforms integrated | PM Kisan, UIDAI, eSanjeevani, e-Shram, Indian Railways |
| App availability | Android and iOS |
| Maha Kumbh 2025 deployment | 11 languages for 450 million+ visitors |
Background and Mission
India has 22 Scheduled Languages (Eighth Schedule of the Constitution) and hundreds of dialects. Despite rapid digitalisation, most digital content and government services were available only in English and Hindi, creating a massive language barrier for:
- Citizens accessing e-governance services
- Students in regional-medium schools accessing quality content
- Gram Sabha proceedings needing multilingual documentation
- Emergency services communicating with diverse populations
The NLTM was conceived to address this through AI-powered language technology as a shared public resource.
AI Capabilities
| Technology | Function |
|---|---|
| Automatic Speech Recognition (ASR) | Converts spoken language to text in 22 languages |
| Machine Translation (MT) | Translates text between Indian languages and English |
| Text-to-Speech (TTS) | Converts text to natural-sounding speech |
| Speech-to-Speech Translation | Real-time voice translation across languages |
| Optical Character Recognition (OCR) | Reads text from images and documents |
| Transliteration | Converts script from one language to another |
Key Deployments and Use Cases
Maha Kumbh 2025 (January-February 2025)
Bhashini was deployed at scale during Maha Kumbh 2025 in Prayagraj, providing multilingual access in 11 Indian languages to an estimated 450 million+ visitors:
- Kumbh Sah’AI’yak Chatbot: AI-powered, multilingual, voice-enabled chatbot launched by PM Modi, powered by Llama LLM
- Digital Lost & Found: Voice-based item registration in native languages with real-time translation
- Emergency Services: CONVERSE feature enabled communication with UP Police 112 helpline in multiple languages
- Real-time announcements: Event schedules and safety guidelines translated into 11 languages
Government Platform Integration
| Platform | Bhashini Integration |
|---|---|
| PM Kisan | Multilingual access for farmer beneficiaries |
| UIDAI (Aadhaar) | Services in regional languages |
| eSanjeevani | Telemedicine consultations in local languages |
| e-Shram | Portal upgraded to all 22 Scheduled Languages |
| Indian Railways | Passenger information services |
| Department of Defence Production | Website in 22 languages (January 2025) |
SabhaSaar — Gram Sabha Digitisation
SabhaSaar converts audio and video recordings of Gram Sabha meetings into structured, multilingual digital records. Since its national rollout in 2025:
- Over 15.6 crore interactions processed
- Averaging 5 lakh+ daily engagements
- Deployed across 2.5 lakh+ Panchayats
Technical Architecture
- Open-source approach: Bhashini models are available as APIs for developers and startups
- Crowdsourced data: Citizens contribute voice samples and translations to improve AI models via the Bhashini app (BhashaDaan)
- Model marketplace: Developers can access pre-trained models for building language-aware applications
- Edge deployment: Models optimised for low-bandwidth environments (rural India)
About Digital India Corporation (DIC)
- Section 8 (not-for-profit) Company under MeitY
- Formerly known as Media Lab Asia
- Manages multiple Digital India initiatives including Bhashini, DigiLocker, and UMANG
- DIBD (Digital India BHASHINI Division) is a dedicated division within DIC for language AI
International Recognition
- PM Modi offered Bhashini technology to SCO (Shanghai Cooperation Organisation) member countries
- Recognised by UNICEF for making digital services accessible through language AI
- Model for other developing nations facing similar multilingual digital access challenges
Latest Developments
- Migrated to sovereign Indian cloud infrastructure (February 2026) — Bhashini moved from a global hyperscaler to Yotta Government Community Cloud and Shakti Cloud (NVIDIA H100-enabled), keeping all datasets, models and user interactions within India’s jurisdiction
- VoicERA launched at the India AI Impact Summit 2026 — an open-source, end-to-end Voice AI stack built on Bhashini infrastructure, strengthening India’s DPI for multilingual voice technologies
- Scale milestone: Bhashini now powers over 500 government websites, processes over 15 million inferences daily, and has surpassed 6 billion total inferences with sub-second response times
- Language support expanded to over 36 languages in text and 22+ languages in voice, up from the initial 22 Scheduled Languages
- Maha Kumbh 2025 deployment validated at scale — Kumbh Sah’AI’yak chatbot (powered by Llama LLM) served 450 million+ visitors in 11 languages; architecture stress-tested on Yotta Shakti Cloud
- BharatGen AI — India’s first government-funded multimodal large language model supporting 22 Indian languages — built using Bhashini infrastructure and domestic datasets
- Department of Defence Production website made available in all 22 Scheduled Languages using Bhashini integration (January 2025)
Prelims Importance
- Full name: BHASHa Interface for India | Launched: July 2022 under NLTM
- Implementing body: Digital India BHASHINI Division (DIBD) under Digital India Corporation (DIC), a Section 8 Company of MeitY
- Languages: All 22 Scheduled Languages (Eighth Schedule) + tribal languages
- AI capabilities: ASR, Machine Translation, TTS, Speech-to-Speech, OCR, Transliteration
- Budget: ₹495.51 crore approved under NLTM
- SabhaSaar: Gram Sabha digitisation tool; 15.6 crore+ interactions since 2025 rollout
- Maha Kumbh 2025: Deployed in 11 languages; Kumbh Sah’AI’yak chatbot powered by Llama LLM
- Government integration: PM Kisan, UIDAI, eSanjeevani, e-Shram, Indian Railways
- International: PM Modi offered Bhashini to SCO countries
- App: Available on Android and iOS; BhashaDaan for crowdsourced language data
Mains & Interview Importance
GS3 — Science & Technology:
- Analyse Bhashini as a Digital Public Infrastructure (DPI) for language. How does it compare with India’s other DPI successes like Aadhaar, UPI, and DigiLocker?
- Discuss the role of AI and NLP in bridging India’s linguistic digital divide. What are the accuracy and bias challenges in multilingual AI models?
GS2 — Governance:
- How can Bhashini transform e-governance delivery in a linguistically diverse country like India? Evaluate with reference to SabhaSaar and eSanjeevani.
- Examine the potential of Bhashini in making the Right to Information and grievance redressal systems accessible in regional languages.
GS2 — International Relations:
- PM Modi offered Bhashini to SCO member nations. Discuss language AI as an instrument of India’s digital diplomacy and soft power.
Interview angles:
- “India has 22 Scheduled Languages but hundreds of dialects. Can Bhashini realistically cover all of them?”
- “Bhashini uses crowdsourced data from citizens. What are the privacy and data quality concerns?”
- “Compare Bhashini with Google Translate. What makes a government-built language AI platform necessary?”
Essay connection: Technology and linguistic diversity, digital inclusion, AI for governance, India as a DPI pioneer, multilingualism as a national asset