India’s Sarvam AI is emerging as a strong contender against ChatGPT and Google Gemini, focusing on regional language support that caters to India’s diverse digital landscape. Launched in February 2026, this Bengaluru-based startup aims to democratize AI access for Indian users with its unique offerings. Its multimodal Large Language Model, Sarvam RAINBOW, and the Sarvam Vision model excel in Optical Character Recognition (OCR), outperforming global competitors in understanding Indian languages and document formats. With scores of 84.3% on olmOCR-Bench and over 90% on OmniDocBench v1.5, Sarvam demonstrates superior accuracy tailored to local needs. Additionally, the Bulbul V3 model provides fluent text-to-speech synthesis across various Indian languages and accents. By specializing in region-specific AI tasks rather than generalist functions, Sarvam AI aims to deliver more relevant, locally impactful solutions, enhancing its position in the AI market. This innovative approach sets Sarvam AI apart in the competitive landscape of global AI technologies.
Source link
Share
Read more