Google has added a serious speed boost to the Gemini lineup: Gemini 3 Flash was officially introduced on December 17, 2025 and has already started rolling out as the default model in the Gemini app, as well as in Search via AI Mode (as it becomes available to users).
The idea behind Flash is straightforward: keep “frontier-level” intelligence, multimodality and tool use, but make it faster and more cost-efficient for everyday work — quick answers, interactive explanations, prototyping, vibe coding and agentic workflows without the feeling that every prompt costs as much as a small rocket launch.
What matters is that this isn’t just a consumer update. Google is deploying Gemini 3 Flash across its enterprise and developer stack too: it’s available in Vertex AI (with docs that cover parameters such as thinking_level) and has already landed in Gemini CLI, clearly positioning it as a new “speed standard” for production and development. Google Cloud Documentation+2
