Google unveils Gemini 2.5 Pro Experimental, its most advanced AI yet, topping the LMArena leaderboard across coding, math, science and more, thanks to a 1-million-token context window. OpenAI integrates image generation into GPT-4o & Sora for seamless conversation and visual interpretation, impacting ChatGPT’s functionality across all plans. Figure’s humanoid robots now walk like humans with refined gait via reinforcement learning and physics simulations – enabling smoother movement and speed, while Alibaba’s Qwen team unveils two new AI models, a versatile one for real-time processing of text, audio, images & video (Qwen2.5-Omni-7B), and another designed for advanced image reasoning across blueprints and geometry (QVQ-Max). Ideogram 3.0 takes the lead in image generation with complex layouts, logos and typography becoming easier to create, while Reve Image 1.0 emerges as #1 on Image Arena with superior prompt adherence, detailed long-text rendering & natural language editing. Microsoft introduces AI agents for workplace efficiency with Researcher and Analyst capabilities, Perplexity refines AI search with Answer Modes and Transactions, and Tencent’s Hunyuan T1 sets a new standard in AI reasoning, outperforming DeepSeek R1, OpenAI’s o1 & GPT-4.5. DeepSeek V3-0324 delivers high performance on PCs using the Mixture of Experts (MoE) approach, while a water-based, shape-shifting robot created by South Korean researchers opens doors for future applications in environmental cleanup and beyond. Lastly, Kepler Robotics unveils Forerunner K2 with state-of-the-art NimbleMaster hands that can lift up to 15 kg, providing greater dexterity in handling objects. These advancements demonstrate the rapid progress of AI, robotics & automation, shaping our technological landscape in unprecedented ways.