What is IPFS?

Imagine looking for a document not by the shelf where you left it, but by a unique identifier that describes its content. That’s the basic idea behind IPFS: you ask for specific data and the network looks for the computers that can give it to you.

On the regular web, a link takes you to a specific location, usually a server. IPFS uses a Content Identifier (CID), which is based on cryptographic hashes. If a file’s content changes, its CID changes too. That way you can verify that the data you receive matches what you asked for.

Files move between nodes, meaning computers taking part in a peer-to-peer network. Not every node needs to store everything. You can also open content through a gateway, without running your own node.

Why it matters

In crypto, IPFS is used for NFT images and metadata, as well as for files of decentralized apps. Instead of storing an entire image on the blockchain, a reference to its content on IPFS can be stored there. IPFS isn’t a blockchain and doesn’t have a native token of its own.

We find it useful to distinguish between a file’s identity and the computer that hosts it. If copies are available on other nodes, the content can stay accessible even if the original provider stops serving it.

What to watch out for

IPFS doesn’t guarantee permanent storage. Someone has to keep storing and serving the data. Pinning is the setting that makes a node keep specific content, instead of possibly deleting it when it clears its cache.

If you’re looking at an NFT, the label “stored on IPFS” doesn’t by itself tell you who maintains the files. IPFS also doesn’t encrypt content automatically: anyone who knows the CID can try to retrieve it. And a correct CID confirms that the data matches, not that the file is safe or its creator trustworthy.