18 February 2026·PIB

VoicERA Launched on BHASHINI National Infrastructure at India AI Impact Summit 2026

In this note

1. At a Glance

VoicERA is an open-source, end-to-end Voice AI stack launched on the BHASHINI National Language Infrastructure as a "national execution layer" for multilingual Voice and Language AI [1][2].
Launched at the India AI Impact Summit 2026 by the Digital India BHASHINI Division (DIBD) under Digital India Corporation (DIC), MeitY [1][3].
Examinable as a flagship of India's Digital Public Infrastructure (DPI) stack and the IndiaAI Mission — relevant to GS-II (Governance) and GS-III (S&T, IT) [1].

Launched on 18 February 2026 at the India AI Impact Summit 2026 in New Delhi [1][4].
Marks BHASHINI's expansion from text/translation services to real-time speech, conversational AI and multilingual telephony at population scale [1][2].

BHASHINI (Bhasha Interface for India) — MeitY's National Language Translation Mission, launched July 2022 by PM Modi, to break the language barrier in digital services [2].
Operates under Digital India BHASHINI Division (DIBD) of Digital India Corporation (DIC), MeitY [1].
BHASHINI hosts 300+ AI-based language models covering ASR, MT, TTS, OCR, transliteration, language detection [2].
VoicERA is the latest addition — the voice "execution stack" sitting atop BHASHINI [1][2].

Name: VoicERA — open-source end-to-end Voice AI stack [1].
Launch date: 18 February 2026 [1].
Venue/Event: India AI Impact Summit 2026 [1].
Parent Ministry: Ministry of Electronics & Information Technology (MeitY) [1].
Implementing body: Digital India BHASHINI Division (DIBD) under Digital India Corporation (DIC) [1].
Launch led by: Shri Amitabh Nag, CEO, Digital India BHASHINI Division [1].
Collaborators: EkStep Foundation, COSS (Centre for Open Source Software), IIIT Bengaluru, and AI4Bharat [1][3].
Design attributes: open, pluggable, interoperable, cloud-deployable, on-premise ready [1].
Core tech: high-fidelity ASR (Automatic Speech Recognition) + TTS (Text-to-Speech) in a modular framework [2].
Coverage ambition: nuances of 700+ Indian dialects [2].

Scientific / Technological

Provides execution layer atop BHASHINI's 300+ models — converts AI assets into citizen-facing voice services [1][2].
Modular, pluggable architecture enables integration with existing govt APIs and service delivery systems [2].
Open-source release lowers entry barrier for startups and academia building Voice AI [1].

Administrative / Governance

Enables departments to onboard voice-enabled citizen services rapidly across agriculture advisories, education, livelihood, grievance redressal, scheme discovery, citizen feedback [1].
On-premise readiness addresses data-sovereignty needs of sensitive departments [1].

Social

Targets the digital divide for non-English, non-literate citizens via vernacular voice access [2].
Coverage of 700+ dialects aligns with constitutional ethos of the Eighth Schedule languages and beyond [2].

Strategic / Sovereign AI

Reinforces India's Sovereign AI posture — locally built, open-source alternative to foreign proprietary voice stacks [1][2].
Part of the broader IndiaAI Mission ecosystem (BharatGen, AIKosha, AI4Bharat) [2].

Economic

Open-source DPI model attracts startup innovation, reduces voice-AI cost for MSMEs and govt vendors [1].

18 Feb 2026 — VoicERA launched at India AI Impact Summit 2026 [1].
2026 — BHASHINI expanded model catalogue beyond 300 AI language models [2].
India AI Impact Summit 2026 — successor in the AI summit series after the UK (2023), Seoul (2024), Paris (2025) AI summits [4].

VoicERA is an open-source, end-to-end Voice AI stack — NOT a proprietary product [1].
Launched on 18 February 2026 at India AI Impact Summit 2026 [1].
Deployed on BHASHINI National Language Infrastructure [1].
Implementing entity: Digital India BHASHINI Division (DIBD) under Digital India Corporation (DIC), MeitY [1].
Launch led by Amitabh Nag, CEO, DIBD [1].
Collaborators: EkStep Foundation, COSS, IIIT Bengaluru, AI4Bharat [1].
VoicERA is cloud-deployable AND on-premise ready [1].
Targets 700+ Indian dialects [2].
BHASHINI hosts 300+ AI language models covering ASR/MT/TTS/OCR/transliteration [2].
BHASHINI = Bhasha Interface for India, launched 2022 under National Language Translation Mission [2].
VoicERA forms the "national execution layer" for multilingual Voice and Language AI [1].
Application sectors flagged: agriculture advisories, education, livelihood, grievance redressal, scheme discovery [1].

GS-II: Government policies/interventions; e-governance; welfare delivery via DPI.
GS-III: Science & Technology — Indigenous AI; awareness in IT; Digital India.
Plausible question stems: 1. "Discuss how India's Digital Public Infrastructure approach is being extended to AI through initiatives like BHASHINI and VoicERA." (GS-III) 2. "Multilingual Voice AI is essential for inclusive digital governance in India. Examine." (GS-II) 3. "Evaluate the role of open-source AI stacks in advancing India's sovereign AI ambitions." (GS-III)

BHASHINI / National Language Translation Mission — parent platform.
IndiaAI Mission (2024, ₹10,371.92 cr) — umbrella AI policy.
AI4Bharat (IIT Madras) — collaborator; foundational Indic NLP models.
BharatGen — India's indigenous multimodal LLM initiative.
Digital India Programme (2015) — overarching framework.
Eighth Schedule languages — constitutional linkage to language policy.
Global AI Summits (Bletchley 2023, Seoul 2024, Paris 2025, India 2026) — diplomacy angle.
Data Sovereignty / DPDP Act 2023 — data protection backbone for voice data.

Wrong ministry: VoicERA is under MeitY, NOT Ministry of Communications or Ministry of Culture [1].
VoicERA is not an LLM/chatbot — it is the Voice AI execution stack atop BHASHINI [1][2].
BHASHINI was launched in 2022, not 2023; do not confuse with IndiaAI Mission (2024).
DIBD sits under Digital India Corporation, not directly under MeitY's main secretariat [1].
AI4Bharat is incubated at IIT Madras, while COSS is at IIIT Bangalore — distinct partners [1].