Google’s Gemini artificial‑intelligence offering has moved from experiment to commercial engine in the last year, triggering a sharp rise in cloud demand that could reshape Google Cloud’s revenue trajectory. Usage of the Gemini API more than doubled over the past year to roughly 85 billion calls, while Gemini Enterprise has grown to about 8 million subscriptions, signaling broadening adoption among businesses that want to embed large‑model capabilities into their products and workflows.
The commercialisation of Gemini is creating a virtuous — and expensive — cycle: heavy model use translates into persistent demand for compute, storage and networking, the core products Google Cloud sells. Customers that commit budgets to AI projects typically require large, ongoing amounts of GPU and CPU time, managed services, data pipelines and security integrations, and many also buy into complementary Google offerings such as Workspace, BigQuery and identity services, increasing overall account value.
The surge matters because cloud providers compete not just with software but with infrastructure economics. Rising API calls mean more metered compute consumption and higher cloud bills, which can translate quickly into meaningful top‑line gains for Google Cloud if margins are preserved. The pattern also strengthens customer lock‑in: enterprises standardising on Gemini for model services are likelier to keep related workloads and data on Google’s infrastructure rather than fragmenting them across rivals.
This growth comes amid an industry‑wide boom in demand for AI‑optimised servers and chips. Hyperscalers and enterprises are racing to provision more GPU capacity to serve real‑time and batch AI workloads, benefitting chip vendors and data‑centre suppliers and intensifying competition among cloud platforms — chiefly AWS, Microsoft Azure, and Google Cloud — to supply the most efficient, cost‑effective stack for generative AI.
Risks remain. Running and fine‑tuning large models is capital‑intensive, and Google absorbs substantial operational cost to deliver low‑latency, high‑availability services. Margin pressure could intensify if price competition forces discounts or if customers migrate models on‑premises. There are also potential regulatory and privacy headwinds as enterprises shoehorn more sensitive data into third‑party models and as governments scrutinise the market power of major cloud vendors.
For global enterprises and cloud rivals, the takeaway is clear: Gemini’s commercial traction is not only a product success but a strategic lever that can deepen Google’s enterprise relationships and lift demand for cloud infrastructure. How far this accelerant will push Google Cloud’s market share depends on Google’s ability to scale capacity efficiently, keep pricing competitive, and reassure customers on security and regulatory compliance.