India’s AI startup Sarvam AI has announced a sharp reduction in pricing for its document intelligence platform, Sarvam Vision, after witnessing strong adoption from developers and enterprise partners. The company says more than 35 million pages are now being digitised using the Sarvam Vision API, prompting it to pass efficiency gains directly to customers through a substantial price cut.
Sarvam Vision was introduced in February as a vision-language model designed for document intelligence tasks. Since its launch, the platform has been used to process large volumes of documents, helping organisations convert physical records into searchable digital formats.
Sharing the update on X, Sarvam AI said, “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 at scale. We’re now passing these gains on by reducing the Sarvam Vision API price from 1.5 to 0.5 per page.”
The revised pricing brings the cost down by nearly 67 percent, making the service considerably more affordable for businesses and developers working with large document collections.
According to the company, the reduction was made possible by improvements behind the scenes as usage of the platform continued to grow. Sarvam said that increasing demand pushed it to optimise several aspects of its infrastructure to handle higher document volumes more efficiently.
Explaining the changes, the company added, “As adoption grew, we began serving users with considerably higher document volumes. So we reworked parts of our serving stack – optimised inference kernels for state-space architecture, smarter page-level batching, better hardware utilisation across our sovereign cloud. The result is a model that runs far more efficiently at scale. We’re passing these gains directly to our users through this price drop.”
Sarvam Vision focuses on document intelligence and optical character recognition (OCR), helping users extract and understand information from scanned files, forms, archives, manuscripts, financial records and other document types. The model has been built with a strong focus on India’s multilingual requirements and supports all 22 official Indian languages.
The startup has previously highlighted the model’s performance on benchmark tests. According to information shared by the company, Sarvam Vision achieved 84.3 percent accuracy on olmOCR-Bench and 93.28 percent on OmniDocBench v1.5. The company says the model is capable of handling scanned documents, complex page layouts and a wide variety of content formats.
Sarvam AI has positioned itself as a sovereign AI company, with an emphasis on building foundational AI technologies tailored for India’s needs while operating on domestic infrastructure. The company’s goal is to make advanced AI tools more accessible to Indian developers, businesses and institutions.
The latest price reduction comes at a time when demand for AI-powered document processing is rising across sectors such as finance, healthcare, education and government services. By lowering costs while scaling usage, Sarvam appears to be betting that wider adoption will further strengthen its position in India’s growing AI ecosystem.
– Ends
Source link
[ad_3]
