IPFS Stats / Explorer?

Curious, what is the total number of nodes running IPFS?
How many transactions per day roughly, overall?

  • Is there an IPFS explorer?

Thanks.

1 Like

About the “main network”, there is no centralised place to check (yet), so it’s a hard question:


The IPFS may have an estimate since there are running very popular nodes (Bootstrap nodes, Preload Nodes, Relay Nodes, Gateway Crawler, etc.)

Plus, IPFS nodes can be used to function in a private network (only talking to nodes in the same private network), or in a local network so it’s hard to count them in any way.

There is no “transaction” per se, because IPFS is not a blockchain or a marketplace. It’s a network of nodes storing and exchanging data.
There is a curently-building blockchain on top of IPFS that will provide incentive for that: Filecoin.

There is no “Explorer” as in “Ethereum Explorer” because there is no consensus on a common “list of files available” (or any common “big file” shared by everyone, such as a ledger in the blockchain world).

There could be a search-engine like Google and some others did for the Web, but they are still very WIP or rudimentary.
Exemple: ipfs-search.com , https://www.ipse.io

1 Like

Thanks. IPFS search is interesting.
Number of nodes that touched IPFS at 120K seems a bit high? Is there online proof of any sort by any chance?

What would constitute a definitive proof? 120k is an estimate from a core developer. You either have to trust him or explore the network yourself.

It would be good if Juan or others corroborated that number.

Can’t just take a number like that.

The source is:


The post is written by Stephen Allen (AKA Stebalien), who is a core developer working full-time for Protocol Labs on Go IPFS and libp2p.

If you trust him:
-On the optimistic side

  • He is talking about 120k nodes using their Relay Nodes. These nodes are nodes that need external help to get out of their NAT. The number of all nodes should be greater.
  • Publishing content is longer and longer on IPFS despite work on performance, which suggests that the network has kept growing since

-On the pessimistic side

  • It might have been a stress test by an attacker (but the fact that it happened after introducing autoRelay and not after introducing Relays suggest it’s organic because this test could have happened before)
  • You could argue that IPFS team have interest in pumping their number, even though they have nothing to sell per se. But it seems you trust Juan so I guess you would trust his core team.

Maybe you can reach out to Pinata.cloud , which should have a good view on the network because they try to be connected to a lot of peers to provide a fast “layer 2 network” as they call it (see blog). If you do, I’m sure reporting back here will be beneficial for a lot of people. :slight_smile:

Can you give examples of sources you would deem acceptable?

@wmougayar The network have kept growing to a few hundreds of thousands, apparently (watch after 1 min):