India Enters AI Race: Sarvam AI will give competition to ChatGPT and Gemini, know why and how it will increase tension?

There is now a growing need in India to create its own AI models to compete with big tech companies like OpenAI and Google. Now, a company claims that it has not only achieved this, but has also left the world’s biggest companies behind. The credit for this success goes to Sarvam AI, at least according to the company’s CEO, Pratyush Kumar, who shared a test chart to prove his claim.

His post said that the company’s AI chatbot Sarvam Vision, which competes with ChatGPT and Gemini, has outperformed the models of big AI companies by performing better in some cases. According to the post, the locally developed AI model scored more than 93% in the OmniDocBench v1.5 benchmark test.

Is India’s Sarvam AI better than ChatGPT?

This benchmark test is used to test the capabilities of an AI chatbot. It measures their ability to recognize complex fonts, hand-written text, and other types of written data. Sarvam AI scored 93.28 percent in this test, which is much higher than the scores of ChatGPT and Gemini.

This is certainly considered a major success, because even though Servum AI was not created by resources like OpenAI and Google, it can handle some difficult tasks better than even the biggest and fastest models on the market. So, what exactly is Sarvam Vision, and how is it outperforming the bigger companies? But the real question is: what is the essence of Sarvam Vision, and how is it outperforming larger companies? It is very important to understand the context behind these tests.

This “Made in India” AI model appears to have been trained on Indian languages ​​and scripts, which contributes to its better performance and response. Google and OpenAI are good, but their scope is very big. With over 3 trillion parameters, these AI models are capable of handling a wide variety of tasks. Based on 3 billion parameters, Sarvam AI is designed to handle visual language models with a special focus on Indian languages. It can understand images and describe what it sees.

Although its capabilities may be limited to certain areas, Sarvam AI and its text-to-speech model, Bulbul V3, clearly show that Indian startups are moving forward quickly. More Indian companies may enter this field in future.