IndiaAI Mission Supports Avataar’s ‘Varya’ Distilled Video Model

IndiaAI Mission launches the indigenous ‘Varya’ distilled video model developed by Avataar with IndiaAI Mission support. The model is a distillation technique to make frontier video AI affordable, accessible, and relevant.

Varya Distilled Video Model (IndiaAI Mission-supported):

Dimension	Key Details
Developer and Support	The model is developed by Avataar with support from IndiaAI Mission.
Type	Varya comprises a distilled video model built to make frontier video AI affordable, accessible, and relevant.
Core Technique	Varya uses a distillation technique that reduces video generation from 50 steps to 4 steps, while maintaining comparable output quality.
Definition	Distilled video generation is a model compression technique where a compact “student” model replicates outputs of a larger, slower “teacher” model by transferring capabilities while eliminating redundant computation.
Cost Metric	The model can generate video at ₹0.48 per second.
Cost Efficiency Claim	The model provides for up to 10x higher cost efficiency.
Product Experience Flow	The product experience is designed around: Idea, Video, and Story.