Meet NVIDIA Personal AI Router Pair, introduced as a free, open-source software tool designed to squeeze extra mileage out of your idle PC computing power for local AI tasks. The tool automatically sniffs out compatible PCs on your local network and shunts individual inference requests over to systems with spare capacity, all while working smoothly with Ollama and LM Studio. CyberLink's new PhotoDirector AI PC mode is a prime example of diffusion models crashing into creative software, launching optimized for NVIDIA RTX Spark within PhotoDirector 365. This mode lets you choose between local or cloud processing for AI-driven editing tools like generative editing, image enhancement, and object removal.

Back in August, the local AI scene was buzzing with major announcements. Nemotron 3.5 Lightning—a 30-billion parameter model capable of running on NVIDIA RTX PCs, RTX PRO workstations, DGX Spark, and Jetson—was already up for grabs. On top of that, a whole parade of AI models and tech updates broke cover, including Z.ai's GLM-5.3-Flash, Qwen's Qwen3.8-Flash-Next and Qwen3.8-27B, LTX's LTX 2.5, MiniMax-H3, Meta's Muse Glimmer, and DeepSeek v4 Flash. These models are heavily focused on keeping local AI workloads humming across NVIDIA's various platforms.