A data syncing layer
that just works.

Iroh is a protocol for syncing bytes.
It's open source, peer-2-peer, and works on your phone.

Join the Beta

01

Purpose-built for efficiency.

We have been working on and around IPFS nearly since its inception, and are starting out on a new implementation because we have unfinished business with IPFS.

We believe the best days of the protocol are ahead, but to get there we need to ship an order-of-magnitude improvement that unlocks new platforms and use cases.

We're building iroh to see just how far we can take IPFS.

Faster

Iroh consistently outperforms Kubo, using fewer resources to serve more requests.

There is a ring of truth to the "rewrite it in Rust" cliché. Adding in lessons learned while working with IPFS over the years doesn't hurt either.

Throughput
6,926
4,957
Adding Content
199 Mb/s
294 Mb/s
Memory Consumption
12,456
1,234
Avg CPU Load
12,456
1,234
Avg time to First Byte
12,456
1,234

Practical

Iroh does fewer things than kubo, with more polish. Iroh targets a subset of the kubo API, aiming at the most commonly-used features of IPFS, while steering users away from pitfalls that have been trapped in kubo to maintain backward compatibility.

If it's shipped in iroh, it will just work.

Platform Specific

Running a single binary in the cloud doesn't make much sense. Neither does duplicating files locally when running IPFS on the desktop

Iroh is a single codebase with multiple platform targets, allowing iroh to be microservices in the cloud, use more natural defaults on desktop.

Mobile Support

Finally, IPFS on a phone.

Iroh can be embedded into iOS & Android applications to speak the IPFS protocol natively. It's not a device-overheating afterthought, it's a ground-up rethink of what IPFS on a phone should look like.