IndiaAI-backed Varya bets on efficiency over size in video AI race

New Delhi: Avataar.ai on Friday launched Varya, an indigenous video-generation AI model developed with support from the IndiaAI Mission, claiming it can generate videos at a fraction of the cost of several leading global models.The company said Varya uses a machine-learning technique called distillation to reduce the number of computational steps required to generate a video from more than 50 to four, lowering computing costs while delivering similar output quality.The model was unveiled in New Delhi in the presence of MeitY secretary S Krishnan. Avataar was among the companies selected under the IndiaAI Mission to develop indigenous AI capabilities using subsidised national computing infrastructure.Several technology companies have recently launched AI systems capable of generating videos from text prompts or images, but the computing cost involved remains high. Avataar said Varya can generate video at about ₹0.48 per second and is up to ten times more cost-efficient than several international models, based on its internal benchmarks.The company said distillation allows a smaller AI model to produce results similar to those of a much larger model while using far less computing power.

According to Avataar, conventional video-generation systems typically require more than 50 iterative steps to produce a final output, whereas Varya completes the process in four steps.The company said the model was trained to generate visuals reflecting Indian settings, including festivals, clothing, food, public spaces and regional contexts. It said the technology could be used by teachers to create visual lessons, by small businesses to develop advertisements and by government agencies to communicate information through videos.Krishnan said, "The launch of one of the foundational models supported under the IndiaAI Mission marks a significant milestone in India's AI journey."He said the initiative reflected the government's commitment to building indigenous AI capabilities and supporting innovation through foundational AI models.Avataar CEO and co-founder Sravanth Aluru said India's AI opportunity would be shaped not only by the largest models but also by AI models that can be used widely at lower cost."For a country of 1.4 billion people, affordability is not a feature, it is a prerequisite," he said.Users can enter a text prompt or upload an image to generate a video and extend it through additional clips. The company said the technology could be applied to lessons, advertisements, informational content and other forms of visual storytelling.Avataar said it would publish a technical report detailing Varya's architecture, distillation methodology and performance benchmarks.Varya is among the first AI foundation models announced under the IndiaAI Mission support programme, which aims to encourage the development of indigenous AI technologies and applications tailored to Indian requirements.