Privacy Technology

What Is Freenet (Hyphanet)?

It predates Tor by years, works nothing like it, and recently went through a rename messy enough to confuse its own community.

Freenet, now called Hyphanet, is a decentralized, peer-to-peer network created in 1999 for anonymous, censorship-resistant publishing and file sharing. Unlike Tor, it doesn't provide access to the ordinary internet — it's a fully self-contained network where users store encrypted pieces of files and 'freesites' across each other's computers, with no central server to target or shut down. In 2023, the original project was renamed Hyphanet after a separate, newer, Rust-based rewrite (previously codenamed Locutus) took over the 'Freenet' name for a different project with different design goals, less centrally focused on anonymity.

Before Tor existed, before 'the dark web' was a phrase anyone used casually, a university student named Ian Clarke published a paper describing a fully anonymous, decentralized way to store and retrieve information — no central server, no single point anyone could shut down.

That project, called Freenet, quietly kept running for over two decades. Then, in 2023, something unusual happened: the project split, a newer, unrelated rewrite took over the original name, and the actual 1999-era network had to rename itself entirely just to keep being findable. Even longtime users needed a moment to catch up.

Illustration of a distributed peer-to-peer network with data fragments scattered across many connected nodes
Founded1999, by Ian Clarke
Current name (classic network)Hyphanet (renamed 2023)
'Freenet' now refers toA separate, newer, unrelated rewrite
Access to ordinary internetNo — fully self-contained network

TL;DR

Quick answer

Freenet, now called Hyphanet, is a decentralized peer-to-peer network founded in 1999 for anonymous, censorship-resistant publishing, entirely separate from the ordinary internet — in 2023, the original network was renamed Hyphanet after a newer, unrelated Rust-based project took over the 'Freenet' name with different, less anonymity-focused goals.

Last reviewed2026-07-27
Reading time9 min read
DifficultyIntermediate
EvidenceStrong
Founded1999, one of the oldest anonymity networks
StructureFully decentralized, peer-to-peer
Reaches ordinary websites?No, self-contained network only
2023 renameOriginal project became Hyphanet
SpeedNotoriously slower than Tor or I2P

The Basics

The network, and the naming mess, explained separately

Freenet, as originally conceived in 1999, is a fully decentralized peer-to-peer network built specifically for anonymous publishing and censorship-resistant file sharing. Rather than connecting to any central server, each participant's computer contributes a portion of its disk space, called a datastore, which holds small encrypted pieces of files and 'freesites' — websites that exist only within the network itself, similar in spirit to Tor's .onion sites but architecturally very different.

When you request a file or freesite, the network retrieves it by piecing together encrypted fragments distributed across many different users' datastores, similar in concept to BitTorrent but built specifically so that no single node necessarily knows what content it's actually storing. This design makes the network genuinely difficult to censor or shut down, since there's no central point of failure, but it also makes it noticeably slower and more bandwidth-intensive than more centralized-feeling alternatives like Tor.

In March 2023, this original network went through a significant identity crisis. A separate, newer project — a ground-up rewrite in the Rust programming language, previously developed under the codename 'Locutus' — was rebranded to take over the 'Freenet' name, with a different core mission: becoming more of a general-purpose 'decentralized web' platform rather than an anonymity-first network. To avoid confusion, the original 1999-era project renamed itself Hyphanet, a nod to the underground fungal networks that connect and share resources between trees.

The two names untangled

  • Hyphanet: the original, classic network from 1999, focused specifically on anonymity and censorship resistance
  • Freenet (2023 and onward): a separate, newer, Rust-based project aiming for a broader 'decentralized web' platform, with anonymity as a lesser priority
  • The two are not interchangeable, don't share infrastructure, and serve meaningfully different goals

How a file request actually moves through Hyphanet

The path a request takes across a fully decentralized network with no central server.

Diagram showing a file request being pieced together from encrypted fragments across multiple peer nodes
1

Request sent

A user requests a specific freesite or file by its unique key

2

Fragments located

The request routes across peer nodes, each holding small encrypted pieces of content

3

Pieces reassembled

Retrieved fragments are decrypted and reassembled locally into the complete file

4

No central server

At no point does any single server hold or control the complete content

Your computer might be storing pieces of files you've never seen and can't identify

When you join Hyphanet, part of the deal is that your device stores small, encrypted fragments of other users' content, without you knowing what that content actually is — you're contributing storage space to a network-wide jigsaw puzzle where you can't see the picture on your own pieces.

It's a genuinely unusual trust arrangement compared to most networks: you're not just anonymous as a user, your own hardware becomes an anonymous, unwitting participant in storing content for the entire network, which is precisely what makes the system resistant to being shut down by any single action.

Misconception

Freenet and Hyphanet are just two names for the exact same thing.

Reality

As of the 2023 rebrand, they refer to two separate, unrelated projects. Hyphanet is the continuation of the original 1999 network. 'Freenet' now refers to a newer, Rust-based rewrite with different design priorities, less centrally focused on anonymity, and the two do not share infrastructure or content.

The mechanics of a fully decentralized anonymous network

The underlying logic is a genuinely different approach from Tor's relay-based routing.

Joining the network

Installing Hyphanet configures your device to contribute a portion of disk space, the datastore, and connect to other participating nodes.

Like joining a community lending library where you also agree to shelve a few unlabeled books for other patrons.

Content gets fragmented and encrypted

When someone publishes a file or freesite, it's broken into small pieces, encrypted, and distributed across many different users' datastores.

A puzzle scattered across dozens of different rooms, with no single room holding the full picture.

Requesting content

To retrieve a freesite or file, a user sends a request using its unique cryptographic key, which the network routes across nodes likely to hold relevant fragments.

Asking around a large library system for scattered pages of the same manuscript.

Reassembly

Once enough fragments are located, they're decrypted and reassembled locally on the requesting user's device into the complete file or page.

Piecing together a jigsaw puzzle once all the separately stored pieces have finally been gathered.

Optional darknet mode

Users can configure Hyphanet to only connect to people they personally know and trust, rather than the wider public network, making detection significantly harder at the cost of reduced reach.

Choosing to only lend books to close friends rather than the general public.

A project built to resist centralized control got fought over by its own decentralized community

There's a fitting irony in the 2023 dispute: a network specifically engineered around the idea that no single authority should control information ended up embroiled in a very centralized-feeling disagreement over who actually had the right to control its own name, with the original maintainers ultimately having to rebrand rather than the newer project choosing a different one.

If Hyphanet is older and more anonymity-focused, why did the new project get to keep the 'Freenet' name?

Given that Hyphanet is the direct continuation of the original 1999 project, why does the newer, less anonymity-focused rewrite hold the more recognizable name today?

According to public accounts of the split, Ian Clarke, the original creator and architect of Freenet, was also the one driving the new Rust-based rewrite, and as the project's original founder, he made the call to apply the established 'Freenet' name to his newer direction, despite disagreement from the maintainers who continued developing the classic network. It reflects a fairly common, if contentious, pattern in open-source projects: the person most associated with founding something doesn't always agree with the community that later maintains it about which direction, or which name, represents its true continuation.

The 2001 paper that started it all

Ian Clarke's original 2001 academic paper, 'Freenet: A Distributed Anonymous Information Storage and Retrieval System,' became one of the most frequently cited computer science papers of 2002, laying out the fragmented, encrypted storage design years before Tor's onion routing or similar anonymity networks reached comparable public attention.

Long before 'the dark web' became a mainstream phrase, this specific research was already treated as a serious, influential contribution within computer science — a reminder that Hyphanet's underlying ideas have academic roots stretching back further than most of what people now associate with dark web technology.

Three networks, three different goals

Laid out directly, the design priorities diverge sharply.

HyphanetTorFreenet (2023, new)
Primary goalAnonymous, censorship-resistant publishingAnonymously reaching the ordinary internetGeneral decentralized web platform
Access to regular websitesNoYes, as a core featureNot the primary focus
Network structureFully peer-to-peer, no relaysVolunteer-run relay networkPeer-to-peer, database-oriented
Anonymity as top priorityYesYesNo, secondary to platform features
mostlyTrue

Is Hyphanet still relevant today, or has it been overtaken by Tor and I2P?

Still relevant for its specific niche — fully decentralized, censorship-resistant publishing — though it remains a smaller, slower, more specialized tool than Tor.

Hyphanet's lack of central infrastructure gives it genuine censorship-resistance advantages Tor's relay-based model doesn't fully replicate, but its slower speed and smaller user base have kept it a niche tool rather than a mainstream Tor alternative.

What this history says about the long, tangled lineage of anonymity technology

Hyphanet's story is a useful reminder that today's well-known privacy tools didn't spring up independently — they sit on top of decades of overlapping, sometimes competing research and design philosophy, much of it developed years before the dark web became a household phrase. The 2023 split, messy as it was, is really just the latest chapter in a much longer pattern: ideas about anonymity and decentralization keep getting rediscovered, forked, and rebuilt by new generations, sometimes even by the very same person who built the original.

Questions people ask

If this got you curious, go here next

What is I2P and how is it different from Tor?

Another alternative anonymity network, with its own different design philosophy.

Is Tor untraceable?

How the most well-known anonymity network compares in terms of real-world limits.

Why the dark web

The broader origin story of anonymity technology, including networks like this one.

Can I access the dark web without Tor?

A direct look at whether networks like Hyphanet count as a genuine Tor alternative.

What is Tails OS and why do people use it?

A privacy-focused operating system built around similar anonymity principles.

The name changed. The idea it was built to defend didn't

Hyphanet's rename was messy, and the confusion with 'the new Freenet' will probably linger for years. But underneath the naming dispute sits the same idea Ian Clarke wrote about in 1999: information that no single authority can quietly delete, stored not in one place, but in thousands of small, anonymous pieces at once.

You now know

  • Freenet, now called Hyphanet, is a decentralized, peer-to-peer anonymity network founded in 1999
  • It's fully self-contained and doesn't provide access to the ordinary internet, unlike Tor
  • In 2023, the original network was renamed Hyphanet after a separate, newer project took over the 'Freenet' name
  • The two current projects — Hyphanet and the 2023 Freenet — have different goals and don't share infrastructure

Common myth

Myth vs reality

Myth

Freenet and Hyphanet are the same thing under two names.

Reality

Since 2023, they refer to two separate, unrelated projects with different goals.

FAQs

Questions people ask

Sources

Further reading

  • Original Freenet academic paper (2001)Ian Clarke and collaborators
  • Official Hyphanet and Freenet project documentation on the 2023 rebrandHyphanet.org and Freenet.org

Continue learning

Next useful step

Keep going

The next door is usually the interesting one

The answer you came for touches a few neighboring questions. These are the ones most likely to make the picture click.

What you should remember

The name changed. The idea it was built to defend didn't

  • Hyphanet's rename was messy, and the confusion with 'the new Freenet' will probably linger for years. But underneath it sits the same idea Ian Clarke wrote about in 1999.
  • Freenet, now called Hyphanet, is a decentralized, peer-to-peer anonymity network founded in 1999
  • It's fully self-contained and doesn't provide access to the ordinary internet, unlike Tor
  • In 2023, the original network was renamed Hyphanet after a separate, newer project took over the 'Freenet' name

A few useful next steps

Where this question wanders next

The dark web is less a single tunnel than a set of side passages. These are the useful ones from here.

If this made you wonder

Freenet collection

Build the basics

1

What Is I2P and How Is It Different From Tor?

Tor gets all the attention. I2P quietly does something structurally different — and understanding the gap matters.

2

What Is 'Tor Browser Dark Web'?

It's a search term that mashes two different things together — a piece of software, and a category of website it happens to let you reach. Untangling them clears up almost every follow-up question.

3

What Is Tor Browser Used For?

Ask people what Tor Browser is for, and most will guess the dark web. Ask the Tor Project's own usage data, and you get a considerably longer, more ordinary list.

4

What Are Deep Web Websites?

Deep web websites are just websites that require a login or aren't indexed by search engines. You use them all the time.

5

What Are Onion Sites, Or On The Dark Web?

Websites that exist entirely within Tor's own network, reachable only through Tor Browser and a very different kind of address.

Questions people ask first

Choose by the time in your pocket