NVIDIA Personal AI Router (PAIR) is software that connects compatible macOS, Windows, and Linux systems with NVIDIA RTX™GPUs and DGX Spark systems into a personal home AI cluster. PAIR distributes local AI inference workloads across available devices while keeping prompts, files, and agent context on the user’s home network.
A personal home AI cluster connects multiple compatible devices on the same local network, allowing them to share available computing resources for local AI inference. PAIR enables Macs, NVIDIA RTX systems, and DGX Spark systems to provide more capacity for local AI applications and agents. The devices remain separate systems that handle parallel tasks; PAIR doesn’t combine them into one virtual GPU.
A node is a compatible system that has been paired with NVIDIA PAIR and made available to the local AI cluster. Nodes can include supported Macs, NVIDIA RTX PCs or workstations, and DGX Spark systems.
Yes. NVIDIA PAIR can pair supported macOS systems with compatible NVIDIA RTX PCs, workstations, and DGX Spark systems on the same local network. This enables Mac users to route local AI inference workloads to available NVIDIA compute resources.
Yes. PAIR discovers compatible NVIDIA RTX and DGX Spark systems on the same local network and helps them operate as one personal AI inference cluster. Workloads can use available compute across multiple devices instead of relying on a single machine.
Yes. NVIDIA PAIR supports compatible macOS, Windows, and Linux systems. Supported hardware, operating system versions, and device configurations may vary, so review the current system requirements before installation.
NVIDIA PAIR requires compatible macOS, Windows, or Linux systems and supported hardware. Requirements may include specific operating system versions, NVIDIA RTX GPUs, GPU memory, system memory, processors, storage, and local network connectivity. See the System Requirements section on the NVIDIA PAIR landing page for current details.
Yes. A supported Mac can be added as a node alongside NVIDIA RTX PCs, workstations, and DGX Spark systems. PAIR coordinates the paired devices for local AI workflows while keeping prompts, files, and agent context on the local network.
PAIR routes inference requests to available devices within the local cluster. When one device is busy or reaches its available capacity, PAIR can direct requests to other compatible nodes, helping reduce bottlenecks during demanding AI and agent workflows.
Yes. PAIR supports local AI applications, agents, and multi-agent workflows that generate multiple or sequential inference requests. By routing requests across paired systems, PAIR helps make more local compute available for demanding tasks.
Yes. PAIR supports Ollama and LM Studio as local inference backends. Applications can use a local endpoint while PAIR proxies inference requests to available compute within the personal AI cluster.
No major application or agent changes are required. PAIR provides a single local endpoint and proxies supported Ollama and LM Studio requests, reducing the need to configure applications for each individual device.
PAIR is designed for private local inference. Prompts, files, and agent context remain on the user’s local network instead of being sent to a cloud inference service.
No. PAIR uses compatible devices already connected to the local network. It does not require special cables, server racks, or traditional data center cluster hardware.
Download PAIR, install it on supported systems, pair the devices on the same local network, and add them to the cluster. Once connected, supported applications and agents can send inference requests through the PAIR local endpoint.
Download NVIDIA PAIR from the official NVIDIA PAIR website. The website also provides current system requirements, installation guidance, and additional FAQs.
PAIR provides a reusable local inference routing layer with device pairing, node discovery, endpoint access, backend proxying, and workload routing. A custom script typically requires developers to manage device availability, endpoints, ports, and application configuration for each workflow.