Google CEO Sundar Pichai called out 'biggest AI budget problem' of companies world over from IO stage with a solution: Companies can save a lot of money by ...

Google CEO Sundar Pichai shifted the AI conversation from to economics at this years’s Google I/O conference. Pichai warned that the companies around the world are blowing through their annual AI budgets by May due to runaway token usage. Pichai said the rapid rise of AI agents has created unprecedented costs for enterprises. “Companies are already blowing through their annual token budgets and it’s only May,” he noted, referring to the billions of tokens consumed by AI systems each month. Pichai said the rapid rise of AI agents has created unprecedented costs for enterprises. “Companies are already blowing through their annual token budgets and it’s only May,” he noted, referring to the billions of tokens consumed by AI systems each month.Pichai unveiled Gemini 3.5 Flash, a model which is designed to rival frontier offerings while cutting costs. He said companies could save over $1 billion annually if they shifted 80% of workloads to a mix of Flash and other frontier models. “If companies used a mix of Flash and other frontier models they could save a lot of money. To put this in perspective, top companies are processing about 1 trillion tokens a day. If they shifted 80% of their workloads from other frontier models to 3.5 Flash, they’d save over $1 billion dollars annually. That is real savings they can pour back into their company,” added Pichai.Google unveiled Gemini 3.5 Flash, its latest AI model, at the I/O 2026 developers conference. The company has positioned this model as the first release in the new Gemini 3.5 family. The company said the AI model is designed to deliver improved performance for coding, agentic AI tasks and multimodal understanding while maintaining the faster response speeds associated with the Flash series.According to Google, Gemini 3.5 Flash is now available globally through the Gemini app, AI Mode in Search, Google AI Studio, Android Studio and enterprise platforms. The model also powers new AI experiences across Google products, including the upcoming Gemini Spark personal AI agent.The company said Gemini 3.5 Flash performs better than Gemini 3.1 Pro across several coding and agentic benchmarks, including Terminal-Bench 2.1, GDPval-AA and MCP Atlas. Google added that the model is capable of handling long-horizon workflows, enabling it to execute multi-step tasks such as application development, code maintenance and document preparation.Google also highlighted improvements in multimodal capabilities, stating that Gemini 3.5 Flash can generate interactive web interfaces, graphics and animations while supporting more complex reasoning tasks. The model is designed to work with Google’s updated Antigravity platform, allowing multiple AI subagents to collaborate on larger workflows.Google’s advantage lies in owning the full stack chips, data centers, cloud, models, and applications. Analysts estimate Google pays 50% to 75% less for AI compute than rivals thanks to its custom TPU chips and direct sourcing. By contrast, competitors like OpenAI rely on Microsoft, Oracle, and Nvidia, paying margins at every layer of infrastructure.