GOOGLE RELEASES GEMINI 3.5 FLASH: FOUR TIMES FASTER THAN ANY RIVAL, AND HALF THE PRICE Google made Gemini 3.5 Flash generally available this week, and the benchmarks are designed to make every competitors pricing team nervous. The model runs four times faster than comparable frontier models, handles a one-million-token context window, and comes in at $1.50 per million input tokens and $9 per million output tokens. For context, that price point undercuts most of what OpenAI and Anthropic currently offer for models with comparable reasoning capability. Flash is not a lightweight stripped-down model. Google designed it specifically for agent tasks and complex coding, the two categories where AI companies are now competing most aggressively for enterprise customers. On Terminal-Bench 2.1, a benchmark measuring how well models complete real-world command-line coding tasks, Flash scored 76.2 percent. It leads in multimodal understanding with 84.2 percent on CharXiv Reasoning. Gemini 3.5 Flash is now generally available through the Gemini API, Google AI Studio, Android Studio, and GitHub Copilot. It is also the new default model powering AI Mode in Google Search globally, meaning more than one billion people are now being served answers by this model every month. The broader signal here is that Google is treating the agent and developer market as the primary battlefield, not the consumer chatbot market. Flash is fast, cheap, and built to run inside other products rather than in front of end users. That is what winning the next phase of the AI race actually looks like. Keywords: Gemini 3.5 Flash, Google AI model, AI agent model, Google I/O 2026