6 prompts to get the most out of Gemini 3, Google's answer to ChatGPT (www.techradar.com)

Google has launched Gemini 3, a major upgrade to its AI lineup that brings sharper reasoning, stronger multimodal understanding and new “generative interfaces” across the Gemini app, Google Search’s AI Mode, NotebookLM and developer platforms. The release includes Gemini 3 Pro for consumer-facing features and a testing-only “Deep Think” mode aimed at tougher, multi-step reasoning. Google says Gemini 3 can handle much longer, messier inputs, break big problems into sub-steps autonomously, and better analyze images and video in context — a step toward an AI layer more tightly woven into everyday apps. Technically the update emphasizes agentic behavior (planning and executing multi-step workflows), dynamic visual layouts (interactive, magazine-style or educational interfaces), and deeper integration with Google services for real-world tasks. TechRadar’s examples show six prompt classes that highlight capabilities: trip planning with tailored itineraries, creating a Van Gogh gallery with context, interactive diagrams for eclipses or plant cells, inbox organization via Gemini Agent, and hands-on logistics like booking a rental car from email. For the AI/ML community this signals a push toward more autonomous, multimodal agents that change prompt design, UX for model-driven apps, and raise questions about safety, permissions and reliable long-form reasoning as these agents gain access to personal data and action-oriented integrations.