Decentralized Inference Framework

An architecture that splits or distributes model inference tasks across multiple independent nodes for execution, rather than relying on a single server. It provides redundancy and resistance to censorship, since no central point controls or can unilaterally block the computation.