Sarvam Vision is finally the sovereign AI model India had been waiting for a very long time. Ever since there has been chatter about AI LLM models in the tech industry only USA and China have taken the spotlight. India not having a locally developed AI model had raised eyebrows despite having a deep talent pool and huge market for AI.

In a series of posts on ‘X’ Sarvam Vision AI co-founder, Pratyush Kumar shared the achievements of the model compared to other dominant models in the market like ChatGPT and Gemini. According to the post by Pratyush Kumar, Sarvam Vision has achieved an accuracy score of 84.3 percent on the olmOCR-Bench. The score is higher than Gemini 3 Pro and recent OCR models such as DeepSeek OCR v2, while ChatGPT ranked significantly lower.

ALSO READ

What is impressive about Sarvam Vision?

The most impressive thing Sarvam AI is how it is beating bigger and dominant models like Chat-GPT, Google’s Gemini and Anthropic Claude on benchmarks of Optical Charachter Recoginition(OCR). This performance benchmark is one of the most test for an AI model, since this is its area of expertise.

In his post on ‘X’ Sarvam Vision co-founder Pratyush Kumar explained how well Sarvam performed in these benchmarks. This has earned praise for the Bengaluru based startup who developed this model from experts and users alike.

What are people saying about Sarvam AI?

Tech commentator Deedy Das said “I was wrong about Sarvam. When I wrote about them a year ago, I felt like the direction to train small Indic language models was wrong. But boy, have they turned it around,” he wrote. “They have the best text-to-speech, speech-to text, and OCR models for Indic languages, and that’s actually really valuable. The pricing is very reasonable.”

ALSO READ

This is despite him underestimating the company for focusing on Indic-language models, is now seeing that scepticism turn into approval.