I2P (Invisible Internet Project): overview, architecture, uses and distinctions

How it works — architecture and key concepts

I2P is an overlay built on top of the public Internet. Participating nodes (routers) form encrypted tunnels and forward traffic for one another. Key concepts include:

• Unidirectional tunnels: inbound and outbound tunnels are created separately so that the path used to send a message and the path used to receive a reply are distinct, improving resistance to correlation.
• Garlic routing: messages are bundled together and encrypted with layered keys so that multiple messages can share the same packet and reduce linkability.
• Router console and proxies: a local router application manages tunnels and peers and exposes SOCKS/HTTP proxies and APIs so user applications can route traffic into I2P without direct knowledge of the routing layer.
• Internal naming and services: websites hosted inside the network (often called eepsites) use the .i2p namespace and are reached using I2P-aware resolution and internal directories.

History and development

I2P emerged in the early 2000s as part of research and volunteer efforts to improve online anonymity. It has been maintained by a distributed community of developers rather than a single corporation and has evolved through incremental improvements: transport protocols, floodfill directory systems for distributing router information, and application integrations. The project favors a practical engineering approach and documentation is community-driven.

Uses and examples

Applications commonly adapted to I2P include anonymous web hosting (eepsites), messaging and chat, blogging, file sharing, and email-like services. Typical usage patterns are:

1. Running an I2P-enabled application that connects to the local I2P router via a proxy or plugin.
2. Hosting an internal site or service intended only for other I2P users.
3. Using distributed services (e.g., forums, peer-to-peer clients that support I2P) to avoid exposing IP addresses to external observers.

Distinctions and notable facts

I2P is often compared to Tor, but there are meaningful differences. I2P is optimized for in-network services and peer-to-peer traffic and uses pseudonymous addressing and unidirectional tunnels; Tor emphasizes low-latency access to the public Internet and uses exit relays. Both systems trade off performance and complexity to achieve different threat models. I2P's design tends to favor persistent, resilient internal services rather than short-term circuits to the clearnet.

Because it is volunteer-operated and distributed, performance and reachability can vary. Users and administrators should understand the operational trade-offs, configure bandwidth and router options appropriately, and consult community documentation when deploying services. For official downloads, documentation, or developer resources, refer to the project pages linked above (project site, source, technical overview).