Avataar AI has launched Varya, a video generation model specifically optimized for the Indian market to provide high-quality content at a fraction of current global costs. For creators and small business owners, this removes the primary financial barrier to using advanced AI tools, making professional-grade video accessible to anyone with a budget.
Varya offers a 20x price advantage over mainstream models. While industry leaders like Veo, Kling, Luma, and Runway typically charge $0.10 or more per second, Avataar plans to price its hosted service at ₹0.48 ($0.005) per second. This shift allows individual creators to generate high-quality clips for much less than previous enterprise-only pricing models.
The model delivers significantly faster processing speeds. By using a distillation technique on the Wan 2.2 model, Avataar reduced the generation process from 50 steps to just 4. On an Nvidia H200 GPU, Varya can produce a five-second 720p clip in 45 seconds, which is 10 times faster than the 1,230 seconds required by the original model.
Cultural awareness is baked into the training data. Unlike many global models that produce generic or stereotyped outputs, Varya was trained on curated data to recognize specific regional nuances. It is designed to accurately identify:
- Local food and clothing styles
- Specific architectural features
- Regional festivals and traditions
Developers can access the model as an open-weight release. Avataar AI will host the model and its training data on India’s AIKosh portal, the government's centralized repository. This means you can self-host or modify the model to fit specific enterprise needs or unique creative workflows without being locked into a single provider's ecosystem.
The launch is backed by a major government initiative. Avataar AI was one of 12 startups selected for the $1.2 billion India AI Mission. This program provides subsidized GPU compute to help Indian developers close the gap with global rivals in model development and deployment.
You can test the tool immediately on their website. Whether you are looking to lower your production overhead or need a model that understands local context, Varya is available now via text prompts or reference images. If you are planning a content rollout for the next quarter, switch to Varya to drastically reduce your per-second generation costs. Read more: Google rolls out Veo 3.1 Lite, halving AI video costs.
