Google Launches Gemini Enterprise for Financial Services with Deutsche Bank, Adds Pay-As-You-Go Pricing

AI agents are making technology spending harder to predict, as costs can vary based on models used, token consumption, and runtime. [1] In response, Google added new pricing options, discounts, and cost-management tools to Gemini Enterprise on Wednesday to help enterprises reduce AI workload costs and improve budget visibility. [1]

The hyperscaler introduced a pay-as-you-go model, allowing enterprises to pay for compute and tokens consumed rather than committing to a base subscription. [1] Additionally, Flexible Savings Plans offer 10% discounts for one-year commitments and 20% discounts for three-year commitments on Gemini Enterprise spending. [1] The pay-as-you-go model lowers the barrier to adoption and is suited to experimentation, temporary projects, and agent workload bursts, according to Stephanie Walter of HyperFRAME Research. [1]

On August 25, Google Cloud launched Gemini Enterprise for Financial Services, an agentic AI platform aimed at credit risk assessment, following a five-year technology rebuild with Deutsche Bank. [2] Deutsche Bank served as a key design partner for the platform's Financial Research agent, which features more than 50 specialized skills and 13 data connectors. [2]

The system can compress a bond portfolio risk exposure analysis into under five minutes and provide automated hedging suggestions. [2] Deutsche Bank's initial deployment is in its Corporate Bank division, with plans to expand into Private Bank and Investment Bank workflows. [2] The platform integrates data from institutional providers such as FactSet, Moody’s, S&P Global, and SEC Edgar. [2]

The dual announcements highlight Google's strategy to address both the operational and financial hurdles of enterprise AI adoption. By releasing industry-specific agents for financial services and legal sectors alongside metered billing, Google is targeting friction points in procurement and vertical integration simultaneously. The financial research agent's ability to pull from institutional data sources like FactSet and Moody's suggests a shift from generic chatbots to deeply integrated enterprise tools capable of complex risk analysis. What the sources don't address: How the pay-as-you-go pricing structure compares in raw cost to the previous per-seat model under heavy, continuous agent workloads.