Launch · AI & Deep Tech

Gnani.ai unveils sovereign AI stack Artha for Indian enterprises

By Startup Enthusiast · 28 August 2026

What they do

Gnani.ai, a voice AI company, launched Artha, an end-to-end sovereign AI stack for Indian enterprises and public institutions

Artha combines Evon v3.3, a 30 Bn-parameter open-weight model trained in 11 Indian languages, and Plexus, an agentic AI platform

The stack aims to help enterprises use AI while maintaining control over their data and technology infrastructure

Gnani.ai unveiled Artha, an end-to-end sovereign AI stack for Indian enterprises and public institutions.
Artha combines Evon v3.3, a 30 Bn-parameter open-weight model trained in 11 Indian languages, and Plexus, an agentic AI platform.
Evon v3.3 provides the underlying language intelligence, while Plexus enables organizations to build and deploy AI agents for specific workflows.
Evon v3.3 is trained on 2 Tn tokens and focuses on reasoning and performance in Indian languages.
Artha was unveiled by India’s vice president CP Radhakrishnan in New Delhi on 28 August 2026.
Evon v3.3 consumes around 40% fewer tokens for Indian-language workloads than comparable models, reducing compute costs.
Gnani.ai plans to expand the model family to 70 Bn- and 100 Bn-parameter models and increase language coverage to 22 languages.
Artha aims to address three challenges: data sovereignty, cost of deploying AI at scale, and multilingual real-world use cases.
Gnani.ai has started showing the model to select customers, with five of 20 enterprises expressing interest and beginning to build on top of Evon.
Evon is expected to contribute 20-30% of the company’s business within the next year.

Gnani.ai was selected under the IndiaAI Mission in 2025 to develop homegrown foundational AI models.
IndiaAI Mission aims to reduce India’s dependence on foreign AI systems by promoting indigenous AI development.
Gnani.ai launched Vachana STT in December 2025, a speech-to-text model for Indian languages trained on more than 1 Mn hours of real-world voice data.
Gnani.ai was founded in 2017 by Ganesh Gopalan and Ananth Nagaraj in Bengaluru.
Gnani.ai raised $10 Mn in a Series B round in 2025 led by Aavishkaar Capital with participation from Info Edge Ventures.

Gnani.ai offers voice AI products for enterprises, including virtual assistants, real-time guidance tools for customer service agents, and voice biometrics for fraud detection.
Gnani.ai caters to enterprises and government bodies across sectors such as banking, insurance, healthcare, telecom, and government services.
Gnani.ai has 200+ enterprise deployments and processes 30 million voice interactions daily.
Gnani.ai has products like Gnani Warp v2.0 (Speech to Speech), Gnani Prisma v2.5 (Speech to Text), and Gnani Timbre v2.5 Beta (Text to Speech).
Gnani.ai has a product called Gnani Evon v3.3, an enterprise-scale language model with deep domain calibration for BFSI, insurance, healthcare, and telecom.

Gnani.ai plans to expand the Evon model family to 70 Bn- and 100 Bn-parameter models.
Gnani.ai aims to increase language coverage from 11 to 22 languages.
Gnani.ai is developing a speech-to-speech model with low-latency reasoning.

Aavishkaar Capital (lead) — Venture capital firm focused on early-stage technology startups in India.
Info Edge Ventures (participated) — Venture capital arm of Info Edge, which owns Naukri.com.

Product: Generative AI voice agents and sovereign AI stack for enterprise automation
Customers: Large enterprises and customer-facing businesses in BFSI, telecom, ecommerce, and healthcare
How it makes money: B2B SaaS providing AI voice and agentic AI platform for enterprise workflows
What sets it apart: Proprietary models trained natively across 11 Indian languages with sovereign data control