Using centralized servers to deliver files and data is known as a client-server model. To put it simply, the client, or user, requests what they want by typing in a URL, and a server delivers it via HTTP.
Decentralized file sharing uses a different model known as peer-to-peer sharing. In the client-server model, the URL points to an IP address, which is based on the server location where the website data is stored.
With P2P sharing, files aren’t stored in a single place. Instead, they’re distributed across a network of many nodes. Each file, or piece of a file, is given a unique cryptographic hash. This means all versions are tracked across the network. When a user requests the data, the network finds the nodes holding a perfect match to the unique hash or hashes.