Cohere launched Command A, an AI model requiring only 2 GPUs that matches GPT-4o in enterprise tasks while processing 23 languages and costing 50% less for private deployments.
Manus AI, developed by Butterfly Effect, operates independently using multi-agent systems to execute complex tasks, though crashes and feedback loops highlight early-stage limitations.
Alibaba’s R1-Omni AI combines visual, audio, and textual analysis with RLVR to achieve state-of-the-art emotion recognition, now available open-source for applications in education, customer service, entertainment etc.
OpenAI has asked the White House for regulatory relief to preempt conflicting state AI regulations, warning that fragmented laws could hinder US innovation and global competitiveness.
Google’s Gemma 3 AI model combines multi-modal capabilities with single-GPU efficiency, offering developers cost-effective access to advanced AI tools. The update includes enhanced image processing, safety filters, and academic support.
Google DeepMind’s Gemini Robotics AI models empower robots to perform tasks like folding paper and packing lunchboxes, leveraging LLMs and safety protocols like ASIMOV.
OpenAI introduced the Responses API and Agents SDK, enabling developers to create AI agents for tasks like web searches, document analysis, and computer automation.
China has set a plan for mandatory AI studies in schools with Beijing piloting the program, expanding nationwide starting fall 2025.
Monica’s Manus AI agent, despite outperforming ChatGPT and Gemini on GAIA benchmarks, faces doubts over limited access, server instability, and a suspended X account. The team admits the product is still in early development.
