Chinese AI firm SenseTime unveiled SenseNova U1, an open-source image model that processes visuals directly and runs on domestically produced chips. The release aims to offset U.S. export restrictions that have limited the company's access to advanced hardware. Available on Hugging Face and GitHub, the model promises faster image generation and interpretation, with support from ten Chinese chipmakers, including Cambricon and Biren Technology. Co‑founder and chief scientist Dahua Lin said the new architecture could help the firm regain ground in the global AI race and accelerate robotics applications. Read more →

Red Hat principal software engineer Sally O’Malley unveiled Tank OS, an open‑source tool that packages the OpenClaw AI agent in a rootless Podman container on Fedora Linux. Designed for power users and IT professionals, the solution isolates credentials, stores state and makes the container bootable, aiming to curb the security risks that have plagued early OpenClaw deployments. O’Malley, a maintainer of the OpenClaw project, said the weekend project was built to give enterprises a safer way to run autonomous AI agents at scale. Read more →

Chinese AI firm DeepSeek released two new large language models, V4‑Pro and V4‑Flash, both featuring a one‑million token context window and open‑source licenses on Hugging Face. V4‑Pro, a 1.6‑trillion‑parameter model, outperformed leading U.S. models in coding and agentic tasks, while V4‑Flash delivered comparable speed at a fraction of the compute cost. DeepSeek also announced a token price of $3.48 per million output tokens, dramatically undercutting OpenAI and Anthropic rates, positioning the models as cost‑effective alternatives for developers. Read more →