Indian AI Startup Targets 500 Million Feature Phone Users With Offline Models

India's Sarvam AI announced plans Tuesday to deploy its artificial intelligence models on Nokia feature phones, automobiles, and proprietary smart glasses, betting that lightweight, offline-capable AI can reach markets that smartphone-centric competitors have overlooked.

The move signals a sharp divergence from the cloud-heavy strategies of OpenAI, Google, and Anthropic. While those companies race to build ever-larger models requiring constant internet connectivity, Sarvam is engineering AI assistants that occupy mere megabytes of storage and run on processors already embedded in low-cost devices.

Speaking at the India AI Impact Summit in New Delhi, Tushar Goswamy, Sarvam's head of Edge AI, said the company has partnered with HMD Global to integrate conversational AI into Nokia and HMD-branded feature phones. A demonstration showed a user pressing a dedicated AI button to query an assistant in a local language about government benefit programs and nearby markets. The company did not specify which features would function without internet access.

"Through edge AI, we want to bring intelligence to every phone, laptop, car, and even a new generation of devices," Goswamy said. He added that Sarvam has collaborated with Qualcomm to optimize its models for the chipmaker's processors, though the company declined to identify specific device targets or deployment timelines.

For finance leaders evaluating AI infrastructure investments, Sarvam's approach presents an interesting counterpoint to the prevailing "bigger is better" orthodoxy. The company's models reportedly run on existing chipsets without requiring expensive GPU clusters or persistent cloud connectivity—a cost structure that could prove attractive for enterprises operating in bandwidth-constrained environments or seeking to minimize ongoing API expenses.

The startup, backed by Lightspeed Venture Partners, Peak XV Partners (formerly Sequoia India), and Khosla Ventures, also revealed partnerships with German industrial conglomerate Bosch to develop in-vehicle AI assistants. Details on that collaboration remain sparse, including whether Bosch plans to integrate Sarvam's technology into its automotive components or merely pilot the system in select vehicles.

Sarvam co-founder Pratyush Kumar unveiled a separate hardware initiative: AI-enabled smart glasses called Sarvam Kaze, designed and manufactured in India. Kumar described the device as a "builders' device" slated for release in May, suggesting the glasses target developers rather than consumers—at least initially.

The announcements arrive as India's government pushes to establish the country as an AI development hub, not merely a market for Western technology. Whether Sarvam's edge-first strategy can compete with cloud-based models remains an open question, particularly as companies like Meta and Google release their own lightweight models optimized for mobile devices.

For CFOs, the calculus is straightforward: if Sarvam's claims hold, edge AI could eliminate recurring cloud inference costs while expanding AI capabilities to users and devices previously considered unreachable. If the models prove too limited for practical applications, the company joins a long list of startups that underestimated the computational requirements of useful AI.

The company did not disclose pricing, revenue projections, or customer commitments for any of the announced partnerships.