Bengaluru-based AI startup Sarvam AI has reduced the price of its Sarvam Vision API by 67% after the platform crossed 35 million pages in document digitisation, driven by strong adoption from developers and enterprise partners.
The revised pricing lowers the cost from ₹1.5 per page to ₹0.5 per page. Launched in February 2026, Sarvam Vision is a vision-language model designed for document intelligence tasks, including optical character recognition (OCR), converting physical records into searchable formats, and handling complex document workflows.
According to the company, rapid usage growth prompted major infrastructure optimisation, including improved inference performance, smarter batching, and better hardware utilisation across its sovereign cloud systems. These upgrades allowed Sarvam to serve higher document volumes more efficiently and pass cost savings directly to users.
Sarvam Vision supports document processing across all 22 official Indian languages and is designed for India’s multilingual and document-heavy environment, handling scanned files, financial records, forms, manuscripts and archives.
Earlier this February, we launched Sarvam Vision, a vision-language model for document intelligence.
Today, more than 35 million pages are being digitised through the Sarvam Vision API by developers and partners.
Since launch, we've made it significantly more efficient to serve… pic.twitter.com/iqjEbZeNGF
— Sarvam (@SarvamAI) May 29, 2026
The company also reported strong benchmark performance, including 84.3% accuracy on olmOCR-Bench and 93.28% on OmniDocBench v1.5.
Sarvam said more than 35 million pages have been digitised through its API since launch, reflecting growing demand across sectors such as finance, healthcare, education and government services. The startup added that lower pricing aims to accelerate large-scale adoption of AI-powered document intelligence tools across India’s digital ecosystem.
