Moonshot AI releases Kimi K3.1 Pro

Google Gemini 4 Carbon wins benchmarks

Google’s new Gemini 4 Carbon model has swept recent benchmark leaderboards, signaling a major push in the race for top-tier AI performance. This release positions Google firmly against competitors like OpenAI and Anthropic by demonstrating state-of-the-art capabilities across multiple evaluation metrics. The move highlights Google's strategy of leveraging its vast research infrastructure to deliver highly competitive foundation models. For developers, this means access to a powerful new API endpoint that may offer improved reasoning and efficiency compared to previous generations. The model’s performance suggests a significant leap in raw capability rather than just incremental tuning. It forces rivals to reassess their own roadmaps for the coming quarter. Access is now open via the standard Google AI API for those ready to test its limits.

llama.cpp b12029 adds CUDA 13 and Snapdragon support

This release quietly cements llama.cpp as the universal inference runtime by finally supporting NVIDIA's latest CUDA 13 stack alongside AMD's ROCm 10.0. The inclusion of native builds for Linux arm64 with Snapdragon chips, covering CPU, Adreno GPU, and Hexagon NPU, signals a serious push into mobile edge deployment. While Apple Silicon KleidiAI is disabled in this specific build, the broader hardware coverage means developers no longer need to compile from source for most modern accelerators. It’s less about new features and more about removing friction for anyone running local models on bleeding-edge or diverse hardware.

Apple Intelligence arrives in China via iOS 27.3