Sarvam Cuts Vision AI Model API Prices by 67% as Document Digitisation Crosses 35 Million Pages

Why It Matters

The steep price reduction lowers the barrier to large‑scale AI document processing, enabling faster deployment of automation solutions and intensifying competition in India’s AI services market.

Key Takeaways

• Pricing cut from ₹1.5 to ₹0.5 per page (~$0.018 to $0.006)
• Over 35 million pages digitised since February launch
• Efficiency gains enable lower costs for developers and enterprises
• Price drop aims to accelerate AI document‑intelligence adoption in India
• Competitors may feel pressure to match Sarvam's new rates

Pulse Analysis

India’s push toward digital transformation has created a surge in demand for AI‑powered document processing. Sarvam AI’s Vision API, a vision‑language model launched in February 2026, quickly became a go‑to solution for converting physical records into searchable digital assets. By handling tasks such as OCR, data extraction, and workflow automation, the platform addresses a critical bottleneck for banks, insurers, and government agencies that still rely on paper‑heavy processes. The company’s rapid scaling—35 million pages digitised in just four months—demonstrates both market appetite and the viability of AI‑driven document intelligence in emerging economies.

The 67% price cut is more than a promotional gesture; it reflects tangible efficiency improvements in Sarvam’s compute and storage stack. Reducing the per‑page cost to ₹0.5 (about $0.006) brings the service into price parity with legacy OCR providers while offering superior AI capabilities. For developers, the lower fee expands the economics of building SaaS products that embed document analysis, turning pilot projects into production‑grade offerings. Enterprises can now justify larger‑scale rollouts, as the cost per digitised page drops below the threshold that previously limited ROI calculations.

Sarvam’s aggressive pricing is likely to reshape the competitive landscape. Global players such as Google Cloud Vision and Amazon Textract may need to reassess their rates in the Indian market to retain customers. Meanwhile, local startups will be compelled to differentiate through niche features or vertical‑specific integrations rather than price alone. Companies planning AI‑enabled digitisation should evaluate Sarvam’s offering not just for cost savings but also for its rapid deployment model, which aligns with the broader trend of moving AI projects from experimentation to enterprise‑wide adoption.