Privacy Jurisdictions Compared

Twelve hosting jurisdictions on one 2026 matrix: alliances, retention, request process, fit. SKN entity plus racks from Zurich to Tokyo; Core from $4.80.
Twelve hosting jurisdictions on one 2026 matrix: alliances, retention, request process, fit. SKN entity plus racks from Zurich to Tokyo; Core from $4.80.
12 locations · 19 coins · SLA 99.95% · no KYC · public canary
VPS from $4.80* — Core on the annual term (list $6/mo), one invoice for 12 months, no auto-charges. Dedicated with IPMI from $39.20/mo. Crypto invoice, no KYC.
Launch nowHow the 2026 matrix was assembled
Jurisdiction pages fail in one of two ways: they drown you in statute citations, or they sell you a flag with a story attached. This matrix does neither. Four columns carry the whole comparison — alliance ties, retention, request process, workload fit — and each cell is either a matter of public record or a published product fact you can check before paying anything.
The sources are named so you can go around this page: FATF guidance on virtual assets frames the payment-privacy axis, the EU GDPR text governs the EU and EEA rows, and the Swiss FADP anchors the Zurich row — all three are linked in the citations block at the end. Alliance classifications follow the standard Five/Nine/Fourteen Eyes lists. Where a claim cannot be sourced to one of those, the cell says less rather than more.
The VPSbit column is one mechanical fact rather than an editorial claim: Meridian Node Ltd., incorporated in Saint Kitts and Nevis, signs the contract and takes the invoice, while each rack answers to the law of the hall it sits in. Twelve sites survived the 2026 catalog cut — Reykjavik, Zurich, Amsterdam, Frankfurt, Luxembourg, Stockholm, Helsinki, Oslo, Singapore, Hong Kong, Tokyo and Taipei — so every row below is a checkout choice, not a wish list.
One deletion deserves its own note: rows for jurisdictions the catalog no longer sells were removed rather than kept for decoration. A matrix that lists flags you cannot deploy to is worse than none, so what follows covers only the twelve racks plus the SKN entity that bills them all. You can verify each row against the locations table in about fifteen minutes, which is exactly how long your threat model deserves.
The 2026 jurisdiction grid
The grid below is the page’s payload: one row per billing entity and per deployable rack, four columns per row, no slogans in any cell. Scan it top to bottom for law, left to right for fit, and note the premium notes in the fit column where they apply — four of the twelve flags reprice, and you should see that before the invoice does.
| Jurisdiction | Alliance ties | Retention theme | Request process | Best fit |
|---|---|---|---|---|
| VPSbit — Saint Kitts & Nevis entity | None | No retention mandate at the entity | Foreign claims litigated locally | Contract venue for every rack |
| Iceland — Reykjavik | None | No mandatory data retention | Speech-friendly tradition | Publishing, archives (+20% premium) |
| Switzerland — Zurich | None | No host-side retention duty; FADP | Judicial oversight | Personal privacy hub (+35% premium) |
| Netherlands — Amsterdam | 9 Eyes | EU GDPR regime | Fast, formalized | Latency, AMS-IX peering |
| Germany — Frankfurt | 14 Eyes | EU GDPR regime | Thorough | DE/CH/AT users, DE-CIX belt |
| Sweden — Stockholm | 14 Eyes | EU GDPR regime, press-freedom tradition | Formalized EU process | Nordic edge, NetNod peering |
| Finland — Helsinki | EU member | EU GDPR regime | Formalized EU process | Baltic/Nordic edge, FICIX |
| Norway — Oslo | 9 Eyes (EEA) | EEA GDPR regime | EEA process | Nordics outside an EU member regime (+20% premium) |
| Luxembourg | EU member | EU GDPR / data-trust regime | Formalized EU process | Quieter EU profile (+15% premium) |
| Singapore | Outside the alliances | PDPA governs | Local legal process | APAC latency, SGIX |
| Hong Kong | Outside the alliances | SAR regime; duties sit locally | Local, complex | Regional APAC, HKIX |
| Japan — Tokyo | Outside the alliances | APPI governs | Local legal process | APAC latency, JPIX/JPNAP |
| Taiwan — Taipei | Outside the alliances | Taiwan PDPA governs | Local legal process | East-Asia edge, TWIX |
Read the alliance column with care, because it is the most abused cell on every jurisdiction page. Membership is not itself a law — it predicts which partners can exchange signals intelligence at low cost, nothing more. The Netherlands staying a 9 Eyes row does not make an Amsterdam rack unsafe for your brochure site; it makes it a labeled latency pick, which is a different and perfectly honest category.
The four columns, decoded
Retention means the records a provider can be forced to store, and it splits into two questions you should never conflate: what local law can demand from a facility, and what this provider keeps as policy. VPSbit’s written position is the same in every location — no payload or guest-traffic logs, billing and panel actions retained, connection metadata kept at most 24 hours — so the retention column above describes the legal environment around that policy, not a different logging practice per flag.
Process measures the friction awaiting anyone who demands your data: warrants, mutual legal assistance, judicial review, appeal windows. Saint Kitts, Iceland and Switzerland raise that friction by lawful design — a foreign claim against an SKN entity litigates in Saint Kitts, and Zurich requests run through judicial oversight under the FADP. Latency picks like the Netherlands and Germany trade that friction for raw speed, which is an honest bargain for public-facing projects and a dishonest one for everything else.
Fit is where you convert the first three columns into a decision, and the premium notes belong to it. Reykjavik (+20%), Zurich (+35%), Oslo (+20%) and Luxembourg (+15%) carry site premiums on the same SKU; the other eight bill at base list, with Core at $6 list / $4.80 annual-eff, Prime at $12 / $9.60, and Titan at $24 / $19.20 wherever you deploy. Law versus latency is a deliberate choice, and the invoice should reflect it consciously rather than by surprise.
One warning about reading any single column in isolation: a “None” in the alliance column does not cancel a weak process row, and strong process does not survive an operator who logs everything. The columns compound, which is why the pairings section next treats the grid as a whole — entity plus rack plus workload — instead of letting you optimize one cell and call it a threat model.
A walking tour of the twelve flags
Start in the north, where the catalog keeps its quietest rows. Reykjavik runs a Tier IV / N+1 hall on entirely geothermal power behind a 10 Gbps uplink on FICIX, under no mandatory data retention — the flagship for speech-adjacent work at a +20% premium. Stockholm answers from a Tier III / 2N site on NetNod and DE-CIX North, 40 Gbps, with a press-freedom tradition inside the EU GDPR frame; Helsinki sits at 20 Gbps on FICIX as the Baltic edge; Oslo, 20 Gbps on NIX, runs EEA GDPR at +20%.
Central Europe is the fast lane. Amsterdam holds a Tier III+ / 2N hall with a 40 Gbps port inside AMS-IX itself — peering so close the exchange is practically a neighbor — under 9 Eyes membership and the EU GDPR regime. Frankfurt rides DE-CIX Frankfurt on 100 Gbps at roughly 8 ms from Amsterdam, thorough in process and unmatched for DE/CH/AT audiences. Luxembourg, Tier IV / 2N+1 at 40 Gbps with LU-CIX and AMS-IX reach, adds a data-trust regime at +15% for workloads wanting EU space with a lower profile.
The alpine row belongs to Zurich alone: Tier IV / 2N+1 on hydro power, 100 Gbps onto SwissIX and DE-CIX, Swiss FADP process outside the EU entirely, judicial oversight on requests — the personal privacy hub at +35%, and the premium most buyers pay without regret once the threat model names it.
Cross to Asia and the catalog keeps its shape. Singapore runs Tier III+ / 2N at 100 Gbps on SGIX under the PDPA; Tokyo matches the 100 Gbps uplink on JPIX and JPNAP under APPI, with a ~30 ms handoff toward Hong Kong; Hong Kong itself is Tier III+ / 2N at 40 Gbps on HKIX, under SAR law with complex local process; Taipei completes the edge at 40 Gbps on TWIX under Taiwan’s PDPA. Eight base-list flags, four premiums, one SKU — the tour ends where it began: law and latency, priced separately.
Pairings that come out ahead
The grid earns its keep when you combine rows, and four pairings cover most real deployments. For speech-adjacent projects, deploy the SKN entity with a Reykjavik rack — no mandatory retention on a geothermal Tier IV / N+1 site, with the contract venue an ocean away from the hall. A personal privacy hub calls for SKN + Zurich: Tier IV / 2N+1 on hydro power behind a 100 Gbps uplink at SwissIX, requests under FADP judicial oversight.
The EU workhorse pairing is SKN + Amsterdam or Frankfurt, where AMS-IX and DE-CIX peering put your European visitors milliseconds from the rack while the entity keeps billing disputes in Saint Kitts. APAC public traffic belongs on Tokyo or Singapore — JPIX and SGIX respectively — when the audience, not the threat model, picks the flag. Each pairing runs on one account with a single checkout, and no pairing requires a second identity file because the first never existed.
| Goal | Primary rack | Replica | Plan cue |
|---|---|---|---|
| Speech-first publishing | Reykjavik | Zurich | Titan $24/mo list |
| EU privacy hub | Zurich | Amsterdam, labeled as latency | Prime $12/mo list |
| Quieter EU profile | Luxembourg | Stockholm | Core $6/mo list |
| APAC serving | Tokyo | Singapore | Titan or Prime |
Facility facts sit under each pairing so you can sanity-check the fit column yourself: Zurich runs Tier IV / 2N+1 at 100 Gbps; Reykjavik pairs Tier IV / N+1 with no mandatory retention mandate; Frankfurt rides the DE-CIX belt at ~8 ms to Amsterdam; Tokyo hands APAC traffic to JPIX at ~30 ms to Hong Kong. The privacy column and the latency column describe one and the same twelve-row catalog, and the SKU you rent is identical in shape everywhere — only the four premium flags reprice it.
A quieter-EU footnote for the third row: Luxembourg’s data-trust regime and Stockholm’s press-freedom tradition give you EU IP space with a lower profile than the AMS-IX/DE-CIX flags, at +15% in Luxembourg’s case and base list in Stockholm’s. Neither is a cloak — both are EU-process rows with the process column filled in — but for workloads that want EU latency without the loudest peering exchanges, they fit.
Entity versus rack: who answers for what
The split at the heart of this page is contractual, not rhetorical. The entity — Meridian Node Ltd. in Saint Kitts and Nevis — handles billing and legal claims: it signs the contract, takes the crypto invoice, and stands as the counterparty if anyone sues. The rack faces hardware seizure: whatever government reaches the building in Reykjavik or Zurich deals with that jurisdiction’s law, its courts and its process, not yours.
You should be precise about what the split does and does not buy. It does mean a foreign plaintiff must litigate contract questions in Saint Kitts, under rules chosen at signup rather than discovered at incident time. It does not mean the rack is beyond the reach of the country it stands in — SKN is the contract venue, not a cloak for the hardware, and any page that tells you otherwise is selling weather. Lawful demands at the facility are answered under facility law, full stop.
For your runbook, the practical consequence is documentation. Note which axis you optimized for each server — law or latency — because a headline cycle, a premium change or a capacity shift can tempt a future you to “promote” a privacy primary into a latency flag during an outage. The two axes are both legitimate; mixing them accidentally is how a deliberate architecture erodes into a default one.
High-stakes deployments deserve one more step before you lean on any of this: read the primary sources — the FATF virtual-assets guidance, the FADP text, the GDPR — and then pay for counsel in the relevant jurisdiction. This matrix compresses the landscape; it does not replace a lawyer who can read your actual threat model and sign their name to the answer.
Dual-homing without a second identity
Dual-homing is routine here, and the catalog is built for it: Zurich or Reykjavik as the privacy primary, Amsterdam or Frankfurt as a labeled latency replica, one account, Monero renewals, and no second KYC file — because there was never a first one to duplicate. The replica exists so a headline cycle or a fiber cut in one hall does not end the project; it is not a place to quietly move the “sensitive” half of the workload.
Keep the replica honest with labels. If Amsterdam exists to serve European visitors fast, the runbook says so next to the server ID, and the privacy primary stays in Zurich. When you actually fail over, the runbook tells future you which box is load-bearing and which axis each one was optimized on — law or latency — so the outage does not become an accidental migration of everything sensitive to the loudest peering exchange in Europe.
Pricing stays predictable across the pairing because the SKU is the same shape at every flag: Core $6 list / $4.80 annual-eff, Prime $12 / $9.60, Titan $24 / $19.20, with quarterly billing at 10% off where a year feels long. The premium flags — Reykjavik +20%, Zurich +35%, Oslo +20%, Luxembourg +15% — are the only repricing you will meet, and they are printed here so the replica’s invoice never surprises you.
Renewals close the loop: pay both boxes on one Monero rhythm and the habit is identical at every flag, which means the privacy primary and the latency replica cost you the same discipline, not just the same dollars. One account, one renewal page, nineteen coins — dual-homing is an architecture decision, not an administrative project.
Law versus latency: running both dials
Most jurisdiction mistakes are not legal errors — they are category errors. A buyer picks a flag for its law, then measures it by its ping, discovers the ping is fine and the law irrelevant to the actual workload, and concludes the whole exercise was theater. The fix is to decide, per server, which dial you are turning, and to write that decision down where the next crisis cannot overwrite it.
The dials genuinely move separately on this catalog. A Zurich primary and an Amsterdam replica can share one account while answering to utterly different regimes — FADP judicial oversight on one side, 9 Eyes adjacency on the other — because the entity signing both invoices is the same SKN company. Twelve rows exist precisely so that nobody has to buy a compromise flag that serves law badly and latency worse.
Run the dials with numbers you measured, not ones you borrowed. Provision the candidate city, time your real workload — the admin SSH path, the webhook round-trip, the database replication lag — and write the figures next to the printed ones: ~8 ms Frankfurt to Amsterdam, ~18 ms Reykjavik to Amsterdam, ~30 ms Tokyo to Hong Kong. Your ISP’s route may disagree with the peering chart, and only your own measurements settle it.
Then guard the assignment. Outages tempt operators to fail everything over to the fast replica and, once the storm passes, to leave it there — a privacy primary quietly demoted into a latency box. The runbook line that names each server’s axis is what stops the drift; the quarterly review is what catches it if the line gets ignored. Law and latency are both legitimate reasons to choose a flag — the only wrong choice is the one nobody wrote down.
Money matters: premiums, terms and renewals
The money side of jurisdiction shopping is short enough to memorize. Four flags carry site premiums — Reykjavik and Oslo at +20%, Zurich at +35%, Luxembourg at +15% — and the other eight locations bill at base list: Core $6, Prime $12, Titan $24 per month. The SKU is the same shape everywhere, so choosing a quieter hall is a premium decision, never a re-quote, and nothing about the pricing depends on who you are.
Terms do the discounting instead. The annual term multiplies any list price by 0.8 — Core lands at $4.80/mo, Prime at $9.60, Titan at $19.20 — as one invoice for twelve months with no auto-charges; quarterly billing takes 10% off where a year feels long. A +35% premium flag on the annual term still prices below the same plan at list in a base hall, which is the arithmetic to run before you dismiss Zurich as expensive.
Renewals are where the privacy design pays its rent. Checkout runs on nineteen coins — XMR and BTC at the front, USDT across TRC-20, ERC-20, BEP-20, SPL, Polygon, Arbitrum and Optimism behind them, plus ETH, SOL, LTC, BNB, TRX, XRP, DOGE, DASH, ADA and GRAM (TON) — and a Monero rhythm keeps the payment leg quiet at every flag equally. One account holds the whole jurisdiction spread, so the renewal page never needs to know you run four countries’ worth of racks.
Audit the money the way you audit the law: once a quarter, list every server with its flag, its term and its premium, and check the total against your runbook’s architecture. Premiums drift into budgets silently; terms lapse into list price when nobody renews deliberately. The point of printed prices is that surprises are optional — treat the invoice as the last column of the matrix, and it will behave like one.
Red flags when shopping jurisdictions
Jurisdiction marketing has its own scam grammar, and it is worth learning before any single-flag host takes your annual prepay. Each pattern below shows up in real sales pages, and each one costs you the exact thing you came here to protect:
- Slogans instead of statutes — “beyond the reach of” claims with no named law, court or process behind them.
- Single-flag privacy hosts, which force one country’s law onto every workload you ever run.
- Providers that can move your rack between legal regimes without telling you, silently re-pricing your threat model.
- Matrices that outlive their catalogs, listing flags the vendor no longer sells — decoration, not deployability.
- Any pitch that promises untouchability rather than published rules; a cloak is a feeling, and feelings do not answer subpoenas.
The self-critical flag belongs here too: jurisdiction matrices decay faster than product pricing. Alliance headlines shift, retention case law moves, and the grid you saved last quarter may already be wrong — which is why this page asks you to reread it at every redeploy and to keep the optimized axis written in your own runbook rather than trusting anyone’s cached copy, including this one.
The honest pitch, restated one last time: an SKN contract entity plus a rack whose facility law you have actually read beats a single-country privacy host every time, because it lets you choose law and latency independently instead of accepting them as one bundle. That is the whole product difference, and it survives contact with the citations at the bottom of this page.
Read before you lean on it
A matrix is a map, and maps deserve a warning label: this one compresses statutes, court lines and alliance politics into single cells, and compression has a price. Before any high-stakes deployment leans on a row above, spend an evening with the primary material — the FATF virtual-assets guidance for the payment axis, the GDPR text for the EU and EEA rows, the Swiss FADP for Zurich — and then buy an hour of counsel in the jurisdiction that will actually answer for your rack.
Three reading threads repay the time. The retention case law out of Luxembourg’s CJEU has shaped what EU authorities can force providers to keep, and it keeps moving; the Icelandic modern-media tradition explains why Reykjavik’s no-mandate row is culture as much as statute; and the FADP’s judicial-oversight design is what makes the Zurich row more than a marketing adjective. Each thread exists because somebody, somewhere, tested the promise in a courtroom.
Reading also calibrates your skepticism, which is the real deliverable. Once you have traced one row from this grid to its source and back, you will never again accept an unsourced jurisdiction claim from a sales page — including anything printed here without a citation. The matrix names its sources precisely so you can disagree with it from evidence rather than from vibes.
Close the loop by dating your own notes. Whatever you conclude about a flag — law, latency, premium, fit — record it with the date and the sources you trusted, because alliance headlines and case law move faster than product pages, and the version of this page you read in October 2026 will not be the last word. Neither will your notes; they will just be yours.
The fast decision path
If you need an answer today, the path is short. Speech-heavy publishing points to Iceland; FADP-minded Europeans to Zurich; a quieter EU IP to Luxembourg; peering-first workloads to Amsterdam, eyes open about the 9 Eyes row; APAC serving to Tokyo or Singapore. SKN signs the invoice in every case, which is what makes each of these a checkout choice rather than a new vendor relationship.
Plan selection rides on the workload, not the flag: Titan at $24/mo list carries speech-heavy publishing and busy APAC serving; Prime at $12 covers the EU privacy hub; Core at $6 — $4.80/mo on the annual term — is the right size for a quieter-EU profile or a first deployment you intend to grow. The recommended pairing on this page is the SKN entity over a Zurich rack: in practice the strongest combination of process friction and facility quality on the list, priced with the +35% premium stated up front.
For the long form behind this matrix, open Best privacy jurisdictions 2026 — the ranking companion that carries the editorial order and its reasoning. The neighboring discipline lives in Best VPS outside the 14 Eyes, Best anonymous VPS for Europeans, Limits of no-KYC anonymity, How to buy an anonymous VPS and Anonymous VPS opsec guide.
The checkout and payments background closes the loop: Minimum-data no-KYC checkout, Buy a VPS with crypto, No-KYC VPS: requirements and limits, Pay with Monero, Offshore VPS, VPS vs dedicated bare-metal, Best no-KYC VPS and FAQ: crypto VPS. One glance at the grid above covers alliance, retention and fit; those pages carry everything around it.
Comparison or ranking — which page should I use?
<p>Fit moves with your workload. The matrix lets you weigh each column against your own threat model, while the 2026 ranking page holds the editorial order and its reasoning. If you already know the axis you care about, the grid answers in a minute; if you want the trade-offs argued, take the long form.</p>
Does a 9 or 14 Eyes flag ever make sense?
<p>For privacy primaries, rarely — that is what the entity split and the non-alliance racks are for. For serving EU visitors quickly, absolutely: Amsterdam and Frankfurt exist in the latency column, and a labeled latency replica beside a Zurich or Reykjavik primary is a deliberate architecture, not a compromise.</p>
Entity location or rack location — which wins?
<p>Each answers for something different: the entity handles billing and legal claims, while the rack faces hardware seizure and facility law. Splitting the two is the whole design, and it is why an SKN-billed Zurich VPS and an SKN-billed Tokyo VPS can share one account and one renewal habit.</p>
SKN entity or Iceland rack — which matters more?
<p>Both, in different roles: the SKN entity is your contract venue wherever you deploy, while the Iceland rack supplies the facility culture — no mandatory retention, speech-friendly tradition, entirely geothermal power. The pairing is stronger than either half, which is why it leads the pairings table.</p>
What do the location premiums change on the invoice?
<p>Four flags reprice the same SKU: Reykjavik and Oslo add +20%, Zurich +35%, and Luxembourg +15%. The other eight locations bill at base list — Core $6, Prime $12, Titan $24 — and the annual term takes 20% off list everywhere, with quarterly billing at 10%.</p>
Where does the deep-dive version live?
<p>The long-form editorial sits on the Best privacy jurisdictions 2026 ranking — this page is its compact matrix companion. Between them you get both speeds: the grid for a decision, the ranking for the reasoning, and the same verdict at the end of each.</p>
Which Nordic flag should I pick?
<p>Helsinki if you want the Baltic edge at base list — 20 Gbps on FICIX under EU GDPR; Stockholm for the press-freedom tradition beside 40 Gbps NetNod and DE-CIX North peering; Oslo if EEA GDPR fits and the +20% premium is acceptable. All three are base-SKU racks with the same entity behind them.</p>
Primary sources
Ready to launch?
Build the box — VPS or bare metal — create the password, pay the invoice that follows.