Open-source routing software that shares AI jobs across devices on your local network.
PAIR is like a mailroom clerk with a wall of lockers. It drops each new box into a slot that is empty.
It helps homes or offices run local AI models on several computers. It sends each request to a suitable device.
Inference engine
PAIR sends requests to available inference engines on the local network.
Ollama
Ollama can be a local inference engine connected to PAIR.
Local-LLM
PAIR lets several devices with local models share different requests.
Routing
PAIR routes requests by device load, online status, and available models.