Tech Desk: India’s AI landscape is about to see a dramatic shift with Avataar AI’s new video model, Varya. Backed by Peak XV and selected for the India AI Mission, Varya isn’t just another video generation tool; it’s a game-changer built specifically for India’s massive scale, offering unparalleled affordability and speed.

This innovative Avataar AI video model can generate a five-second 720p clip in just 45 seconds, a staggering 10 times faster than its base model, Alibaba’s Wan 2.2. More impressively, Varya slashes costs, charging a mere ₹0.48 ($0.005) per second – roughly 20 times cheaper than leading global models. This dramatic price reduction is crucial for making AI video accessible across India, from students and teachers to MSMEs and public services.

Crucially, Varya addresses a common flaw in global AI models: cultural blindness. Trained with curated data, this Avataar AI video model deeply understands Indian cultural nuances, recognizing diverse festivals, food, clothing, and architecture. This ensures outputs are relevant and authentic, avoiding generic or stereotyped content.

Slated for release as an open-weight model on India’s AIKosh portal, Varya will empower developers to self-host and modify it. This initiative aligns with India’s broader AI ambitions, focusing on creating robust applications and a thriving developer ecosystem. With the government targeting $200 billion in AI investment by 2028, Varya represents a significant leap towards an AI-powered India.