Google Gemini 3.5 Flash: Cheaper AI for Businesses

Many large companies have already used up their entire annual budgets for artificial intelligence. Google is addressing this with a new, less expensive AI model. At Google I/O, CEO Sundar Pichai announced Gemini 3.5 Flash. This model is faster and cheaper, designed for enterprise tasks. Pichai noted that top companies on Google Cloud process about one trillion tokens daily. He says companies are burning through their token budgets quickly. The high costs are due to AI agents performing complex, multi-step tasks. These tasks use many resources. For example, Uber used its entire 2026 budget for AI coding tools in just four months. Microsoft even canceled AI coding licenses in one division due to cost. Gemini 3.5 Flash offers strong performance at a lower price. It can match frontier model performance at up to a third of the cost. The idea is to use cheaper options for routine tasks and save premium computing for complex reasoning. This matters because it could save businesses over $1 billion annually.