Sarvam full-stack pivot: Sovereign India chooses vertical integration

Sarvam AI moves from the model to the full stack - model + product + distribution - in a market where public procurement conditions access to data.

The fact

Tech in Asia, July 31, 2026. Sarvam AI, an Indian lab, announces its shift to a full-stack model - proprietary model + end-user product + distribution. The note contextualizes this shift as a response both to investors (perceived medium-term profitability) and to public demand (MeitY is pushing for an Indian sovereign stack, cf. the OpenAI/Anthropic pause in ministries in mid-July).

Under the hood

Full-stack, in India, means one specific thing: model → consumer or vertical app → integration with the public digital layer (Aadhaar, UPI, DigiLocker). This is the pattern that Krutrim and Bhashini have started to follow. The bet: in a market where public demand conditions access to scale data (Indian languages) and where the pause decreed on US providers creates a void, vertical integration is the only way to monetize without depending on a foreign cloud.

So what

Two threads converge: india-ai-sovereignty (the state chooses the stack) and india-ai-native-founders (product reconstruction around the model). Sarvam becomes the practical case - not just a model, a service. To watch: revenue structure (public contracts vs SaaS), presence on Indian language benchmarks (IndicNLPBench), and commercial response from Krutrim / Bhashini.