BIS Research

02/10/2026 | News release | Archived content

India’s Sarvam AI Emerges as a Sovereign Powerhouse, Outperforming ChatGPT and Google Gemini

In a landmark development for India's technology ecosystem, Bengaluru-based startup Sarvam AI is making significant strides with its homegrown sovereign artificial intelligence models challenging global AI giants such as ChatGPT and Google Gemini while redefining how AI can serve India's diverse linguistic landscape.

Sarvam AI, founded in 2023 by Vivek Raghavan and Pratyush Kumar, set out with a clear mission: to build foundational AI systems optimized for Indian languages and cultural contexts, addressing a gap that foreign models often struggle with. The startup's focus on multilingual and voice-centric solutions has positioned it at the forefront of India's sovereign Artificial Intelligence initiative and garnered global attention.

A major breakthrough came with Sarvam Vision, a 3-billion parameter vision-language model designed for document intelligence and optical character recognition (OCR) across English and 22 Indian languages. This language model has demonstrated higher accuracy than Google's Gemini 3 Pro and even OpenAI's GPT systems on specific benchmarks, particularly in reading, understanding, and extracting complex documents with technical tables, formulas, and mixed layouts tasks where many traditional OCR systems falter.

Alongside Vision, Sarvam AI also released Bulbul V3, its flagship text-to-speech (TTS) model, aimed at delivering natural, expressive, production-ready voices across 11 Indian languages. Independent evaluations show that Bulbul V3 excels in naturalness, robustness, and stability across real-world input types, offering a compelling alternative to global voice-AI platforms.

These innovations are not just technological feats but also part of a broader strategic push. Under the IndiaAI Mission, the Government of India selected Sarvam AI from dozens of contenders to build the nation's first sovereign Large Language Model (LLM) a foundation model trained and deployed entirely within India using domestic infrastructure and expertise. This initiative aligns with India's goals of technological self-reliance, data sovereignty, and inclusive digital transformation.

Also Read - Large Language Model (LLM) Report

Sarvam's achievements come at a time when global AI trends show models like Google Gemini dominating search interest in India and capturing widespread adoption. Yet, Sarvam's specialized focus on Indic languages, voice interaction, and document intelligence fills a gap left by larger, more generalized models leading to admiration from industry commentators and users alike.

As the startup continues to refine its technologies and expand capabilities, Sarvam AI's success stands as a testament to India's growing role in shaping next-generation AI not just as a user of global systems, but as a builder of world-class, sovereign models tuned for the country's unique linguistic and cultural needs.

Also read- AI/ML Industry trends

BIS Research published this content on February 10, 2026, and is solely responsible for the information contained herein. Distributed via Public Technologies (PUBT), unedited and unaltered, on March 11, 2026 at 10:39 UTC. If you believe the information included in the content is inaccurate or outdated and requires editing or removal, please contact us at [email protected]