IndiaAI Mission-backed Varya aims to make AI video creation affordable and scalable
Indigenous video-generation model developed by Avataar.ai claims 10x efficiency gains, reducing generation steps from 50 to four
MeitY Secretary S. Krishnan at the launch of Varya, India's first distilled AI video-generation model developed by Avataar under the IndiaAI Mission, in New Delhi on Friday.
New Delhi: India’s push to build indigenous artificial intelligence capabilities received a boost with the launch of Varya, a video story-generating AI model developed by AI transformation company Avataar with support from the IndiaAI Mission. The model is designed to make high-quality AI video creation more affordable, accessible and relevant for India’s diverse use cases.
Unveiled in New Delhi in the presence of MeitY Secretary S. Krishnan, Varya is described as India’s first distilled video model, built to generate culturally contextual visual content across regions, languages, festivals, communities and everyday Indian scenarios. The model is intended for applications ranging from education and public information dissemination to advertising and e-commerce.
According to Avataar, Varya employs a distillation technique that reduces video-generation processes from more than 50 computational steps to just four while maintaining comparable output quality. The company claims the model can generate video at an estimated cost of ₹0.48 per second, making it up to ten times more cost-efficient than several leading global video-generation models.
The platform follows a simple “Idea → Video → Story” workflow, enabling users to generate videos from text prompts or images and extend them into longer visual narratives. The company believes this approach can help teachers, small businesses, content creators and government agencies create video content at scale.
Speaking at the launch, Krishnan said the model represents a significant milestone in India’s AI journey and reflects the country’s commitment to developing indigenous AI capabilities. He noted that support for foundational AI models under the IndiaAI Mission is aimed at fostering research-led innovation and strengthening India’s deep-tech ecosystem.
Avataar CEO and co-founder Sravanth Aluru said India’s AI opportunity would be shaped not only by larger models but also by more efficient and accessible ones. He said affordability is essential for AI adoption at population scale and that the next generation of digital experiences, lessons, advertisements and services could be created through AI-powered storytelling tools.
The company plans to publish a technical report detailing Varya’s architecture, distillation methodology and performance benchmarks.
Distilled video generation is a machine-learning technique in which a smaller model learns to replicate the outputs of a larger model while eliminating redundant computation. Avataar says Varya applies this approach to video generation to deliver faster and more cost-effective performance without significantly compromising quality.
About Avataar
Avataar.ai is an AI-native transformation company focused on building domain-specific AI products that help enterprises improve efficiency, create new operating models and develop intellectual property-led capabilities across high-growth markets.
