Google Unveils Gemini 3.5 Flash for Faster, Low-Cost Enterprise AI Automation
Google used its annual I/O developer conference to unveil Gemini 3.5 Flash, a new artificial intelligence model it says is faster, more capable, and far cheaper to run than the systems many businesses use today. The company is positioning the model as a major shift for enterprise AI, claiming it can handle complex, multi-step work, coding tasks, and agent-style automation while cutting costs enough to save large organizations more than $1 billion a year in some cases.
According to Google, Gemini 3.5 Flash is built to do what older models struggled with: carry out sustained tasks across multiple steps without slowing down or becoming prohibitively expensive. The company says it outperforms Gemini 3.1 Pro on a range of coding and reasoning benchmarks, including tests such as Terminal-Bench 2.1, GDPval-AA, and MCP Atlas, while also performing strongly on multimodal reasoning tasks. Google also says the model produces output about four times faster than competing frontier AI systems, a claim aimed at enterprises looking to reduce latency and operating costs.
The launch reflects Google’s broader bet that the next wave of AI will be driven less by simple chatbots and more by “agents” — systems that can take actions on a user’s behalf. As reported by TechCrunch and VentureBeat, Gemini 3.5 Flash is designed for autonomous and semi-autonomous workflows, from software development and code maintenance to document processing and decision support. Google says the model is particularly effective for long-horizon tasks, meaning work that requires the AI to remember context and act consistently over many steps.
The model is being rolled out broadly across Google’s ecosystem. It is now available in the Gemini app, AI Mode in Search, Google AI Studio, Android Studio, Gemini Enterprise, and Google’s enterprise agent platform, according to Google’s developer and cloud documentation. That broad distribution suggests Google wants Gemini 3.5 Flash to become a default building block for consumer tools, developer workflows, and workplace automation at the same time.
Google also used I/O to introduce related products built around the new model. TechCrunch reported that Gemini Spark, a 24/7 personal assistant that can act on a user’s behalf and integrates with Gmail, is powered by Gemini 3.5 Flash. Separately, Google unveiled Gemini Omni, a multimodal model that can reason across text, images, audio, and video, and can generate or edit video through conversation. The company also introduced updates to YouTube search and Shorts, including an “Ask YouTube” conversational search feature and new Gemini-powered tools for video discovery.
For enterprises, the appeal is not just performance but price. Google says Gemini 3.5 Flash delivers near-Pro-level intelligence at Flash-level cost and speed, a combination that could make AI deployment more practical for companies that have been constrained by inference costs, the expense of running AI models at scale. If the company’s benchmarks and pricing claims hold up in real-world use, the model could lower the barrier for firms that want to embed AI into coding, support, operations, and content workflows without relying on larger, slower, and more expensive systems.
The broader significance of the launch is that Google is trying to turn its Gemini platform into a full AI ecosystem rather than a single chatbot. That strategy puts it in sharper competition with OpenAI and Anthropic, while also giving businesses and developers more options for building agent-based systems. What happens next will depend on how the model performs outside Google’s demos and benchmarks, but the company is clearly betting that speed, automation, and lower costs will define the next phase of AI adoption.