Start here

Frequently asked questions

25 min readView as markdown ↗

Short answers about what URnetwork is, what it stores, and how to use it. The overview explains the full architecture, and the threat model keeps the complete technical record.

The basics

Is URnetwork a VPN?

It does what you use a VPN for: it encrypts your traffic and changes the address websites see. The architecture differs. A VPN concentrates everyone into datacenters with publicly catalogued IP ranges. URnetwork exits through members' real residential connections in 2,000+ real cities across 90+ countries. The path is you → extender → operator → provider → website. On the relayed path the provider never learns who you are, and the apps can seal traffic to the provider once you switch on Post Quantum Encryption, which ships off. The overview explains what each party sees.

Do I have to share my internet to use it?

No. Using the network and providing to it are separate switches. Installing an app never turns your device into an exit for anyone else. Providing is opt-in; see "Sharing your connection" below.

Is it really free?

Yes. The free tier is a daily data allowance, with no card required, and it is not a trial. Paid members (UR Pro) fund the network, and the free allowance refreshes every day out of that. There is no ad or data-sale model. Current allowances and prices: ur.io/products.

Who is behind URnetwork?

BringYour, Inc., a Delaware, USA company, operates the network. Everything is open source at github.com/urnetwork: the apps, the SDK, the protocol engine, and the operator platform. The protocol effort is documented at ur.xyz.

Has URnetwork been audited?

No independent audit covers the protocol, the connect engine, or the operator's server code. That is the layer the privacy claim rests on. Two third-party assessments from 2025 cover other surfaces.

Web application and API: a credentialed, manual penetration test by an independent firm, April to May 2025, against the OWASP Top 10 with a review of OWASP ASVS controls.

Android app: a Leviathan Security Group assessment under Google's App Defense Alliance MASA programme, May 2025, at assurance level AL2. It passed. Leviathan states the work "should not be read as a holistic security evaluation or comprehensive penetration test," and it covers Android only.

Neither assessment examined logging, retention, or the data path. What URnetwork offers instead is continuous verification: all code is public, F-Droid rebuilds the Android app from published source, and egress is a fleet of member devices rather than one custodian.

What platforms can I use?

All of them: Android (Play, F-Droid, Solana dApp Store, direct APK), iPhone and iPad, macOS, Windows, Linux, browser extensions for Chrome, Edge and Firefox, and a web app at ur.io. Every app is open source. Each platform has a getting-started guide in these docs.

Does it cover my whole device, or just my browser?

The OS apps cover the whole device. Android, iOS, macOS, Windows, and Linux tunnel everything the device sends: mail, music, games, updates, not only the browser. The browser extension covers browser traffic. The ur.io web app is the extension's control plane, not a separate product: sign in once on the web and it sets up, connects, and re-points the extension for you.

Can I run URnetwork alongside another VPN?

Not as two device tunnels. An operating system runs one system VPN at a time, so URnetwork and a work VPN cannot both cover the whole device. The browser extension is a proxy, not a VPN, so it can run on top of another VPN and cover browser traffic while the other VPN holds the device.

Privacy

Can URnetwork itself see what I do?

Not once you switch on Post Quantum Encryption, which ships off. The provider sees the destinations it carries but never learns whose traffic it is; the operator stands in between, knowing your account and which provider carried you. With the seal on, the session is encrypted end to end to the provider: the operator cannot read what you send, and no single party sees both who you are and what you do. With it off, the default, the operator relays packets it can read. It parses only the routing header and writes no destination to its ledger, but that is discipline, not encryption.

ModeOperator seesProvider seesDefault and availability
Relayed sealedyour account and source connection, provider association, ciphertext, timing and volumedestination traffic, device and contract id, not your real source IPthe native apps, opt-in: switch on Post Quantum Encryption, which ships off
Relayed standardyour account and source connection, provider association, inner destinations and packet bytesdestination traffic, device and contract id, not your real source IPthe native default, out of the box, since the seal ships off; the browser and proxy paths always; with the seal on, a provider that cannot be sealed to is skipped rather than served here
Directless relay involvementyour real source IP and destination trafficopt-in, by turning off Strong Anonymization

Comparing notes gets them little: the operator writes no destination to its ledger and cannot read one through a sealed session, and the provider holds no identity to contribute.

Limit: provider independence is not measured, and the operator could itself run providers. See the threat model.

What does URnetwork log?

URnetwork's main service records do not contain a browsing history.

It stores: account details you provide, a keyed hash of your IP block, an approximate city, provider associations, and byte counts.

It does not store in the transfer ledger: destinations, hostnames, URLs, SNI, or ports.

Limit: the ingress load balancer's access logs have not been audited, and the production destination of service stderr is not established. See the threat model for the full record.

If police show up with a warrant, what can be handed over?

Only what is stored can be handed over, and no browsing history is stored.

What exists: keyed hashes of address blocks, an approximate city per connection, client ids, byte counts, and your sign-up details. The sign-up details are what can name you: an email or phone number if you gave one, a wallet address, a payment record.

Limit: BringYour is a US company and can be compelled like any operator. What is stored decides the outcome. The threat model itemizes the records.

Who are the providers?

Members. A provider is anyone who runs the app and turns providing on, sharing their connection as an exit. Assume some are curious.

Limit: provider independence is not measured, and nothing prevents the operator from running providers itself. The threat model works through the hostile-provider case.

Can a provider see what I do?

A provider sees what an ISP sees, for the slice of traffic it carries: destination addresses and site names. It does not learn who you are; your real IP stays with the operator unless you opt in to direct mode. It does not read your contents; HTTPS stays end-to-end encrypted to the site.

Can a provider tamper with my traffic?

Not by impersonating a site. Certificate checking happens in your browser, so a provider faking your bank produces a warning, not a working copy. Running many providers gains an attacker little; no provider learns whose traffic it carries. The hostile-provider case is worked through in the threat model.

If everything is HTTPS already, why do I need this?

HTTPS hides what you send. It does not hide who you are or where you went. Your ISP, your employer, and every café network still see the name of each site you connect to, and each site sees an address that identifies your household. URnetwork moves that trail to parties that each hold only a fragment of it. Encryption and anonymity are different problems.

What does my own ISP see?

An encrypted tunnel and nothing inside it. Your ISP sees that your device holds a TLS/QUIC connection to the network, or to an extender that looks like an ordinary service, plus volume and timing. Destinations and contents travel inside the tunnel.

What does "post-quantum" cover?

It seals the session between your app and the provider once you switch it on; it ships off. With it on, the operator in the middle carries bytes it cannot read. The key exchange is X25519MLKEM768, a hybrid of a classical curve and the NIST-standardized ML-KEM, so traffic recorded today cannot be decrypted by a future quantum computer. It covers that session's key exchange; identities stay Ed25519.

Limit: the browser extension and the proxy endpoints have no sealed session, and cannot: their client device runs inside the operator, which translates the browser's or proxy's protocol into the network's. A seal started there would begin at the operator. See the threat model for the full mechanism.

How do I know the sealed session is on?

It starts off, so the first answer is to check the switch. The control is named Post Quantum Encryption and lives in the connect drawer, not in Settings; a fresh install is not sealed until you turn it on. Every current provider build supports the provider side.

The stronger answer is that while it is on you cannot end up unsealed without knowing. The client refuses to send or accept your traffic in the clear, so a provider it cannot establish a session with is skipped rather than used unsealed — you lose that provider, not your encryption. What the app still cannot do is show a per-connection sealed-or-not verdict; that indicator has not been built yet.

Using it

How do I check it's actually working?

Open ur.io/ip. It shows the address and city the internet currently sees for you, whether that address is connected through URnetwork, and both your IPv4 and IPv6. The IPv6 line catches a half-configured setup. Different sites reporting different exits is normal here, not a fault.

Will my bank lock me out?

It shouldn't: your bank sees one steady residential address for the whole session, which raises fewer fraud flags than a shared datacenter IP. The client pins each site to one provider, so the address does not change mid-session. Banks react to sudden country jumps more than to addresses. Pick a location near you, or keep the bank off the tunnel with a split rule.

Why do different sites see different IP addresses for me?

Because your traffic exits through several providers at once, commonly three to eight. Per-site affinity keeps each site on one provider, so any single site sees one steady address. The variety is across sites, not within one.

What do I do when a site blocks me or floods me with CAPTCHAs?

Disconnect and reconnect, which draws a fresh provider window and fresh exit IPs, or pick a different city. Residential addresses attract fewer wholesale blocks than datacenter ranges, but none is pristine: a home IP can carry history from whoever held the lease before. For one stubborn site, add a split rule that routes it around the tunnel.

Can I keep one app, like my banking app, off the tunnel?

Yes. Per-app split tunneling is available on Android and Windows: pick the apps that bypass the tunnel, and everything else stays inside it. Split rules by domain or IP address are available on Android, Apple platforms, Windows, and Linux, so you can exempt one bank's domain without exempting a whole app.

Is there a kill switch?

Yes, on every platform: a toggle in each app on Android, iOS, macOS, Windows, and Linux, plus the extension's own. When it is on and no provider is reachable, the client refuses to fall back to your ordinary local route, so traffic stops. Every toggle starts off, the extension's included. Android adds an OS-level mode that holds even if the app dies (Always-on VPN plus Block connections without VPN).

Limit: on iOS and macOS, enforcement is client-level. Kill the tunnel process outright and the system restores normal networking.

Which location should I pick?

"Best available" picks the highest-scored providers near you and is the right default for speed. Browsing the list shows countries. To land somewhere specific, type a place name and a Cities section appears in the results. The provider count shown per location is live.

Can a provider fake its location?

Not by declaring one. The operator works out where a provider is from the connection it observes, so a provider cannot type in a city it is not in. A location is listed only while a member's device is online in it.

Limit: this is an address-derived estimate. An address that looks like a datacenter, a VPN, or a re-announced range is demoted rather than trusted, not excluded outright.

How fast is it?

URnetwork puts average streaming speed on the network at 40 Mbps+. Speed is a design goal: the path stops at four legs instead of chaining hops, to keep latency down. Your own figure varies, because traffic rides through another member's real connection; their uplink and their distance from you both count. The client replaces slow providers automatically, so a member closing their laptop usually means a brief stall, not a dropped session.

How do I make it faster?

Reconnect, or pick a nearer city. The remaining speed setting is a trade: Strong Anonymization, on by default, is what keeps the operator between you and the provider. Turn it off and the client connects directly to the provider. That is faster, and that provider then sees your real IP address. Make the trade knowingly.

Does it drain my battery?

The battery cost is that of any VPN tunnel. The Android service also backs off under battery-saver and thermal pressure.

Can I torrent through it?

No. BitTorrent is dropped deliberately: a member's home connection would receive the complaint, and no filter can tell a Linux ISO from a movie. Connecting first does not get around the check, because it also runs on your own device. Plaintext BitTorrent is stopped there before it reaches a provider, and encrypted BitTorrent within a flow's first few packets.

Limit: the kill switch decides what happens to retries. With it off, after your device drops an encrypted torrent connection, your torrent app's retries to that peer leave over your own connection for ten minutes, outside the tunnel and with your own address. With it on, they stop.

How is torrenting detected without recording what I do?

By checks at the edges, not by central monitoring. Signatures catch plaintext BitTorrent on any port, and encrypted file-sharing falls to the rule against opaque traffic that matches no recognized protocol. The checks run on the provider's device where traffic exits, past the end of the sealed session, so the operator inspects nothing.

A signature match drops the packet and raises a flag carrying only the peer's device id and a yes/no: no destination, no domain, no contents. The operator has no handler for that flag today, so nothing is stored on arrival. Other drops raise nothing at all. See the threat model.

Can I run an Ethereum node through it?

The execution client, yes, with discovery v4 left on. The security filter drops traffic that looks like random data and matches no recognized protocol, because encrypted file-sharing looks the same. An up-to-date filter recognizes Ethereum's discovery v4 and RLPx peer connections by cryptographic checks on their packets, so both pass. Discovery v5 cannot be recognized, so it is dropped: its packet headers are masked with the receiving node's id, which a provider never sees.

Leave discovery v4 enabled in your execution client. With the kill switch off, your own device sends discovery v5 over your local connection instead, outside the tunnel; with it on, discovery v5 stops.

Limit: a consensus client discovers peers only over discovery v5, so its discovery does not work through the tunnel.

Does URnetwork work where VPNs are blocked?

It is designed to. The extender leg is community-run relays on their own addresses, answering on believable service ports, plus a DNS-shaped transport for networks where only DNS gets out. Blocking the platform's name does not block the way in, because each extender forwards from its own address. The client races direct routes first and falls back to extenders; there is no promise against every censor. An extender cannot read what it carries, and as your first hop it does see your address.

What are the HTTPS/SOCKS5/WireGuard endpoints?

They are bridges for software that cannot run the apps or SDK: a router, a stock WireGuard client, a tool that only speaks SOCKS. The apps remain the most private path. Hosted proxies shows how to use the HTTPS and SOCKS5 endpoints in other apps.

The WireGuard bridge shows the trade. The tunnel address is one the platform allocates, not your home address. But it stays fixed for the life of the profile and is reused with every provider, so providers could link those flows to one client. The bridge has no sealed session and cannot have one: your endpoint of it runs inside the operator, which translates WireGuard into the network's own protocol. One public DNS resolver is hard-coded into the profile. Profile export is a Pro feature, created in your ur.io account. See the overview.

Adblock providers

What is a DNS preset?

A DNS preset is a one-tap choice of who answers your DNS lookups. DNS is the lookup that turns a name like ur.io into an address, and a filtering adblock provider blocks a name by not giving out its real address.

The first pick, URnetwork (default), is URnetwork DNS. It is encrypted DNS over HTTPS through the tunnel, answered by Cloudflare, by Quad9 at its unfiltered address and by DNS.SB. It filters nothing, so ad blocking is off unless you pick it.

The picks sit at the top of the Custom DNS screen, under Traffic DNS, grouped by what they do: Block ads and trackers, Block malware, Family, Use my account and No filtering. Every pick is available on every URnetwork plan, free included, and none is ever chosen for you. The URnetwork ad blocker switch is a separate blocker, built into the app. Both can be on, and the built-in one answers first.

Which adblock providers can I pick?

Six: AdGuard DNS, Cloudflare, Control D, DNS4EU, NextDNS and Quad9, in the groups below. Each pick is named for its adblock provider and what it does, such as AdGuard Family or Cloudflare Malware Blocking. Within a group the picks are listed alphabetically, not ranked, and each one shows what it blocks, where the company behind it is based and a link to its privacy policy.

GroupPickWhat it blocks
Block ads and trackersAdGuard Ad BlockingAds, trackers and phishing
Block ads and trackersControl D Ad BlockingAds, trackers and malware
Block ads and trackersDNS4EU Ad BlockingAds, trackers, malware and phishing
Block malwareCloudflare Malware BlockingMalware, not ads
Block malwareDNS4EU Malware BlockingMalware and phishing, not ads
Block malwareQuad9 Malware BlockingMalware and phishing, not ads
FamilyAdGuard FamilyAds, trackers and adult sites, and it turns on safe search
FamilyCloudflare FamilyMalware and adult sites, not ads
FamilyControl D FamilyAds, trackers, malware, adult sites and drugs
FamilyDNS4EU ChildMalware, adult sites, gambling, drugs and violence, not ads
FamilyDNS4EU Child, no adsAds, trackers, malware, adult sites, gambling, drugs and violence
Use my accountAdGuard Personal, Control D Personal, NextDNS PersonalThe filters you set up in your account there
No filteringAdGuard No Filtering, Cloudflare No Filtering, Control D No FilteringNothing
No filteringDNS4EU No FilteringOnly what courts order it to block
Adblock providerCompany and jurisdictionPrivacy policy
AdGuard DNSAdGuard Software Ltd, Cyprusadguard-dns.io
CloudflareCloudflare, Inc., United Statescloudflare.com
Control DControlD Inc, Canadacontrold.com
DNS4EUWhalebone, s.r.o., Czech RepublicDNS4EU (PDF)
NextDNSNextDNS Inc., United Statesnextdns.io
Quad9Quad9 Foundation, Switzerlandquad9.net

DNS4EU applies court-ordered blocks to every variant, its No filtering pick included. Its resolvers all sit in the EU, so exits elsewhere wait longer for answers.

Below the groups, Custom keeps the full editor of DoH URLs and DNS servers, and it always stays. A filtering server mixed with URnetwork's own servers does not filter, because the fastest answer wins. Last comes Regional, marked unencrypted, which appears only while your exit is in a country where encrypted DNS is known not to work. It uses that country's recommended DNS servers, unencrypted but still through the tunnel, and is never picked for you.

How do I use my NextDNS, Control D or AdGuard DNS account?

Pick Use my account, choose your adblock provider's pick (AdGuard Personal, Control D Personal or NextDNS Personal) and enter your ID. You can paste the bare ID or the DNS address your adblock provider's dashboard shows; the app takes the ID out of it and drops any device name.

Adblock providerWhere to find the IDLooks like
NextDNSthe Setup page of your accountabc123
Control DShow Resolvers on your endpoint, in the dashboardabcd1234
AdGuard DNSyour device's page in the dashboard1a2b3c4d

In your account's settings, keep the block response at zero IP (0.0.0.0, the default) or NXDOMAIN. REFUSED and other responses look to the app like an outage. Blocks answered with zero IP also show in the lookups blocked count.

A wrong ID often filters nothing, because NextDNS and AdGuard DNS answer an ID they do not know without filtering. So once you are connected, the app tests the ID. It tells you if Control D does not recognize it or AdGuard DNS rejects it; otherwise it says it cannot confirm the profile, and never shows a check mark. NextDNS also stops filtering after 300,000 queries a month on its free plan, and AdGuard DNS's free plan has the same monthly cap.

Your ID travels only inside the encrypted DNS request. It stays on your device, is never sent to URnetwork's servers or written to logs, and is deleted when you log out. Only its own field shows it; summaries and status lines name the pick, such as NextDNS Personal. Linking an IP address, which the dashboards suggest, is not needed: your lookups leave from other members' home connections, so linking one would tie a stranger's home address to your account.

What happens when an adblock provider stops answering?

It depends on the group, because a family filter that quietly stopped filtering would defeat its purpose.

  • Block ads and trackers and No filtering: the app switches to URnetwork DNS and says so: "Ad blocking paused: AdGuard Ad Blocking isn't answering". A No filtering pick says "Cloudflare No Filtering isn't answering. Using URnetwork DNS" instead. Your pick returns on its own once the adblock provider answers.
  • Block malware, Family and Use my account: lookups stop resolving until the adblock provider answers again or you switch. Names already in the cache can keep working for up to five minutes.

Either way the status line offers Use URnetwork DNS, which switches in one tap, and Change location, since an adblock provider can be unreachable from one exit and fine from another. Nothing ever falls back to unencrypted DNS or to your local network.

Are lookups filtered while the tunnel connects?

Yes, unless Fast DNS on connect is on. That switch, off by default, answers lookups over your local network while the tunnel connects or reconnects. With an adblock provider picked, URnetwork DNS answers those early lookups, never the adblock provider, so it never sees your own address. Lookups in that short window are not filtered. Once the tunnel is up, your pick answers everything.

What does the "lookups blocked" count mean?

While a pick from any group but No filtering is on, the Custom DNS screen counts the lookups your adblock provider blocked during this connection. It reads, for example, "120 lookups blocked by AdGuard Ad Blocking this connection". The count is kept on your device and never sent anywhere, not even with a support report. It starts again at zero each time you connect or change adblock provider.

It is a floor, so the real number can be higher. The app counts only the blocks it can recognize. A plain "no such name" answer, which some account settings give for blocked names, is not counted, and neither are the blocks of the URnetwork ad blocker.

Who sees my lookups when I pick an adblock provider?

The adblock provider sees each name you look up while connected, and when. Your lookups reach it encrypted, through the tunnel, from the address of your exit, never from your own. With Use my account it also sees your ID, so it can tie those lookups to your account there. Its own privacy policy applies; the table above links each one.

URnetwork's servers learn nothing new. Which adblock provider you picked, your ID and the lookups blocked count all stay on your device.

What can go around my DNS choice?

Anything that does its own DNS. A browser with its own secure DNS setting turned on sends lookups to the resolver it names. So does Android's Private DNS set to a hostname, which the Android app flags on the connect screen. Both skip the URnetwork ad blocker too. Apple's DNS profiles do not apply inside the tunnel. After a switch, apps and browsers can keep earlier answers for about a minute.

Sharing your connection

Can I earn by sharing my connection?

Yes. Turn on providing in the app, or run the provider CLI on a spare machine. Public traffic you carry is metered under signed contracts, and providers participate in the UR protocol; ur.xyz documents how rewards work. You can also restrict providing to your own devices, unpaid.

What does providing cost me?

It costs uplink and ISP data, within limits you control. Providing uses your uplink and counts against your ISP's data cap, so it suits a device on a charger. Carrying other members' traffic never spends your own URnetwork allowance, but traffic between your own devices does. It is set to Wi-Fi only out of the box, so it will not spend your mobile data.

If someone does something illegal through my line, do police come to me?

Sharing is engineered so the traffic that draws complaints never leaves your connection. The connect engine's ip_security layer runs on your own device and drops copyright-class file-sharing and intrusion-class attack traffic at your egress. The categories that produce DMCA notices and abuse reports are removed at the source. Its threat lists are rebuilt from public malware, abuse and botnet feeds with each release, so keeping the app updated keeps the filter current.

The full mechanism is in the threat model, and the terms of service cover providing.

Accounts and billing

How do I pay?

Card via Stripe, the App Store or Google Play in the apps from those stores, or crypto: USDC on Solana, Monero, or shielded Zcash, and under More coins, TAO, Bitcoin, ETH and native SOL. USDC is a dollar-pegged stablecoin; Solana Pay handles it today, and x402 lets an agent pay in-band on any quoted chain. Monero and Zcash go to a one-time address on URnetwork's own nodes, so no payment processor sees them. TAO, Bitcoin, ETH and SOL go to a deposit address made for each invoice; TAO's is a plain Bittensor address, so any Bittensor wallet works. A coin that can't be quoted at the moment is shown greyed out as temporarily unavailable. A crypto payment is one-time: it buys a year of Pro or a data pack and does not renew. Card numbers stay with the payment processor. Ways to pay lists the options on each platform; current prices are on ur.io/products.

Can I pay with TAO from any Bittensor wallet?

Yes. A TAO invoice is paid to a plain Bittensor address, so any Bittensor wallet works. Send plain TAO from the wallet's main balance, not alpha or staked funds. TAO has no payment link, so the invoice shows the address and the exact amount, each with its own Copy button, and its QR code holds only the address: enter the exact amount in your wallet.

How anonymous is payment?

You can pay without attaching a name. One tap creates an email-free Instant Account backed by a recovery phrase, the mechanism numbered-account VPNs use. Wallet sign-in works by signature only; URnetwork never asks for a wallet's seed phrase or private key.

What the payment itself reveals depends on how you pay. Card payments identify you to the payment processor. A payment in USDC, TAO, Bitcoin, ETH or SOL is public by nature: a public ledger leads back to the paying wallet. Monero and shielded Zcash show no sender or wallet to us or to anyone reading the chain.

Whichever you use, we know which account paid, how much and when. We keep the link between a cryptocurrency payment and the account it paid for, for 30 days after the payment is credited. Until then the transaction reference is kept against your account; then the link is removed.

Limit: support cannot recover a secret account. Write the recovery phrase down.

What happens when I hit the free daily limit?

New traffic through URnetwork stops until the free data refreshes at 00:00 UTC. It is a stop, not a slowdown: a transfer runs until the data it reserved is used, then fails rather than crawls. The apps and ur.io/app say that the balance is out and when it refreshes. The apps never disconnect you: your traffic waits in the tunnel, and you disconnect to go back to your ordinary line. Upgrade to Pro for a monthly multi-terabyte allowance on top of the free daily data, or top up with a one-time data pack, no subscription required. About your data covers the details, the browser extension included.

Does traffic between my own devices use my data?

Yes. When one of your devices connects through another device on the same account, that traffic now counts against your account's data balance, like all other traffic. A phone that uses your home computer as its exit is one example. On the free plan it uses your free data; on Pro, your Pro data and the free daily data.

Open connections between your devices show in Pending until they close. If your balance runs out, connections between your devices stop with the usual out-of-balance alert until your balance renews or you add data.

How do I delete my account?

Delete account in the apps, or in your account settings at ur.io; the extension has none, so use ur.io. Deletion removes your account and associated personal information. Cancel an App Store subscription first: deletion is refused while one can renew. Deleting cancels a card or Google Play subscription (manage or cancel).

Troubleshooting

What do I do if it won't connect?

Try three things first.

  1. Toggle the connection off and on. The client draws a fresh provider window.
  2. Check that the OS granted VPN permission. Android: accept the first-run consent dialog in the app, not the quick tile. iOS and macOS: approve the VPN profile in Settings.
  3. On a hostile network, give it a minute while the client rotates through transports and extenders. On a hotel or airport portal, disconnect, sign in to the Wi-Fi, then reconnect.

Each platform's tour in these docs has a full troubleshooting section.

Where do I get help?

Use the Support tab in any app, which files a report and can attach client logs, or feedback.ur.io. One quick triage: if you can sign in at ur.io, your account is fine and the problem is client-side. Security reports go to security@ur.io under the vulnerability disclosure policy.