What Is a Network State? Every Serious Project, Mapped
A network state is an online community that organizes like a startup: shared identity first, then crowdfunded physical territory, then eventually diplomatic recognition from existing countries. Balaji Srinivasan defined the concept in his 2022 book The Network State. As of 2026, no project has completed all three steps. Most are stuck on the first.
Published 2026-06-12 · by Jordan Urbs
The pitch usually arrives breathless: nation states are dying, the cloud is the new country, reserve your citizenship now.
Slow down.
The network state is an interesting idea wrapped in some of the most overheated marketing in the sovereignty world. This guide separates the two — the concept in plain English, then every serious project with its honest 2026 status.
The idea, translated out of the whitepaper
Balaji Srinivasan (former Coinbase CTO, former a16z general partner) published The Network State in July 2022.
His formal definition runs to a single dense paragraph. Translated into normal speech, the sequence has three steps.
Start a community online and build a real shared identity, with a real internal economy. Crowdfund physical territory (scattered pockets around the world, not one connected landmass). Then behave enough like a country, for long enough, that existing countries start treating you like one.
That last step, diplomatic recognition, is the whole claim. Plenty of groups have managed some version of steps one and two… step three is the part nobody has touched.
And the count, as of mid-2026: zero projects have completed the sequence. The network state is a category of attempts.
Four things people mix up
Readers genuinely confuse these, so a clean taxonomy before the map:
Special economic zone (SEZ). A country loosens business rules inside part of its own territory. The government keeps full control; you’re a tenant with better terms. Shenzhen and the Dubai free zones are the famous examples.
Charter city. A host country passes a law letting a city run its own regulations (courts, business law, sometimes taxes) while the host keeps criminal law and foreign policy. More autonomy than an SEZ, still entirely dependent on the host honoring the deal.
Intentional community. People choose to live together around shared values, fully under existing law, with no new governance claim at all. The Free State Project (roughly 6,000 libertarians moved to New Hampshire since 2001) is the largest working example.
Network state. Online-first, founder-led, aiming at eventual recognition. The first three categories exist inside a state’s permission. The fourth claims it will someday graduate beyond permission.
Most projects wearing the network-state label are, on inspection, one of the first three. That’s not an insult… it’s just what the map shows.
The map, project by project
Six projects carry the label seriously enough to track. Each links to its full listing, and the whole set lives in the communities directory.
Próspera: operational, and fighting for its legal life
Próspera is a private charter city on the Honduran island of Roatán, running since 2020 under the country’s ZEDE law. It has real buildings, a few thousand residents and remote e-residents combined, bitcoin accepted as legal tender internally, and a growing medical-tourism niche.
It also has a problem: Honduras repealed the ZEDE law in 2022.
Próspera’s answer is an arbitration claim against the Honduran state for roughly $11 billion (an international lawsuit decided by a tribunal instead of a national court) — and to keep building while the case grinds on.
That detail frames everything below. The most built-out project in the category depends on winning a legal fight with the country it sits in.
Praxis: $525 million raised, nothing built
Praxis was founded in 2019, raised $525 million in 2024, and claims over 150,000 online citizens.
As of mid-2026 it has built nothing. After years of criticism that it was marketing without a city, it announced a California site called Atlas in 2025. The announcement is still all there is.
Treat the citizen count and the timeline with skepticism. Seven years and half a billion dollars in, the deliverable is still a waitlist.
Liberland: 11 years, zero recognitions
Liberland was proclaimed in 2015 on Gornja Siga, a seven-square-kilometer parcel of disputed Danube riverbank between Croatia and Serbia. It issues citizenships, runs blockchain-based governance, and holds gatherings at Ark Village on the Serbian side.
Eleven years in: no UN member state recognizes it, and physical settlement remains minimal.
Liberland is the cleanest test of the recognition thesis, because it skipped straight to claiming territory. The result so far suggests recognition doesn’t follow from declarations, however well-lawyered.
Edge City: real gatherings, no state
Edge City runs month-long popup villages, temporary live-work communities descended from 2023’s Zuzalu experiment. Its flagship, Edge Esmeralda, returned to Sonoma County from May 30 to June 27, 2026, with 500-plus participants, and the organization says a permanent village is coming.
The gatherings are real. The permanent town is still a plan. A popup that everyone leaves after a month makes a great conference format (genuinely), and it is not a country.
Cabin: the neighborhood model
Cabin, founded in 2021, connects 20-plus independently operated coliving neighborhoods across the US, Europe, Latin America, and Asia under shared culture and tooling. It began as a DAO, and the neighborhood-organizing work has since spun out into the Neighborhood Village Project.
Notice the structure: no single company holds the residents. The neighborhoods run themselves. Cabin is smaller than its early hype suggested, and it’s also the only project on this list our directory doesn’t label a trusted third party… because there’s less third party to trust.
Network School: a paid program in a landlord’s towers
Network School is Balaji’s own residential experiment, running since September 2024 in Forest City, Malaysia (a famously underpopulated Chinese-built development near Singapore). Several hundred residents from 70-plus countries combine coliving, coworking, fitness, and a crypto-flavored curriculum, and Malaysian ministers have floated fast-track visas.
Read that plainly. The author of the network-state book is, several years after publishing it, running a paid program inside towers owned by the developer Country Garden, under Malaysian law.
That’s not a gotcha. It may be the honest version of step one. But it’s a school, not a state.
What would have to be true
For any of these to become what the book describes, several hard things need to go right at once:
- A host country honors the deal across elections. Honduras passed the ZEDE law, then a new government repealed it. Próspera’s $11 billion arbitration is what “the deal changed” looks like in practice.
- Or recognition arrives without a host at all. Liberland has run that experiment for 11 years. Current score: zero.
- People actually move. Praxis claims 150,000 citizens and has no physical residents. Network School houses several hundred at a time. The gap between online sign-ups and packed bags is the category’s defining number.
- The founding company survives. Funding runs out, founders pivot, companies get acquired. A country that can be acquired isn’t a country yet.
Will one of them thread all four? I honestly don’t know. Próspera winning its arbitration would be the strongest signal the category has produced… and even that would prove a charter city can defend its contract, which is a different thing from a network minting sovereignty.
The trust question the marketing skips
Run these projects through the same trust test we apply everywhere else in this directory.
Bitcoin in your own wallet is trustless (nobody can stop you, nobody can take it). A network state is the opposite case: a trusted third party with a founder, a balance sheet, and a legal dependency on at least one actual government.
Your residency is a contract with a company. Your “citizenship” is a row in a database that company maintains. Your protections route through Honduran arbitration, Malaysian visa policy, or California land-use law.
None of that makes these projects bad. It makes them services, and sometimes interesting ones.
But “sovereign” branding doesn’t make a thing trustless, which is exactly why every listing here carries a trust label. Five of the six projects above are marked trusted third party, each with a note naming who you’d be trusting. The labels exist because marketing in this category routinely claims the opposite of the structure.
So what do you do with this?
If the idea pulls at you, fellow builders, start by reading each project’s honest status (the communities hub keeps them current), visit the ones that allow visits, and treat any money you send as payment for a service. If the future-country part ever materializes, count that as a bonus you never priced in.
The online half of the thesis looks real. People do organize, fund, and build together across borders now, at meaningful scale.
The state half remains unproven after four years of well-funded attempts. Watch Próspera’s arbitration… that ruling will say more about this category’s future than any launch video.