Google Gemini AI Pro vs Ultra: India pricing reveals agentic roadmap

This pricing deep dive shows how Google is turning Gemini 3 from a generic chatbot into a layered subscription stack built around agents. AI Pro targets prosumers with better context and productivity tools, but AI Ultra is where the roadmap gets interesting: massive NotebookLM quotas and Project Mariner, a browser‑based agent that can autonomously handle research, shopping and travel flows across multiple tabs.

Functionally, that’s a move toward “AI as operating system” rather than “AI as app.” If Mariner and its successors work well, users will increasingly delegate whole tasks—"book me a trip under ₹40,000 with these constraints"—instead of issuing atomic queries. That creates both a moat for Google’s ecosystem and a tighter feedback loop between Gemini, Chrome and Search, blurring the boundaries between them.

In the race to AGI, these tiers also highlight how monetization and capability are becoming intertwined. The most agentic, infrastructure‑heavy features are gated behind AI Ultra, which only some users and enterprises will pay for. That segmentation could shape who gets to push the frontier of real‑world deployment: small teams hacking on AI Pro, or larger organizations buying into Ultra to build deeply integrated agents on top of Gemini.