I’ve been diving into privacy tools lately and stumbled across decentralized VPNs (dVPNs) – those routing traffic through distributed node networks instead of centralized servers. Pretty fascinating concept, but I’d love to hear from folks who’ve actually used them.
A few things I’m trying to figure out:
What dVPN services hold up? Looking at various roundups but would rather hear from actual users
Performance vs traditional VPNs – are we talking meaningful speed drops, connection issues, or are some actually competitive?
Trust model – do you personally feel more confident with a distributed network vs established providers? What’s the real threat model here?
Integration with security-hardened environments – anybody running these alongside Qubes OS or similar setups?
Which dVPN would you recommend for 1. Hardened threat model? 2. Targeted threat model?
I get this is pretty niche territory, but if you’ve tested one and have honest feedback (especially the pitfalls), I’d really appreciate hearing your experience before I start experimenting on my own.
This isn’t my forte. I have to ask what a dVPN solves over Tor and what use case, as I’m also unsure myself. Honestly a dVPN sounds like Tor but managed by an entity to provide more features on top.
ObscuraVPN maybe kind of falls in the category as a 2 party VPN (exit node always differs), but I’m not sure if this counts for your use case.
I’d say it depends on your threat model. I would use tried and true technologies, warts and all, as those warts are understood, with more precarious models. If you don’t have as much to lose, might be interesting to evaluate the new players in the game carefully,
This will strongly depend on what nodes the app or whatever happens to choose for you. Since the nodes are usually run by random people in different hosting services, it’ll depend on whether the server they’ve chosen is powerful enough to handle all the traffic, whether there’s enough bandwidth, etc.
In contrast, a traditional VPN provider controls all the nodes, so they can more easily monitor potentially problematic nodes and upgrade their network if need be.
Anecdotally, I’ve tried Nym, and I’ve found the speeds (with the “fast” WireGuard setting) to vary from somewhat faster than IVPN and Mullvad to much slower. The slower nodes tend to be in countries with few nodes and “good” policies like Iceland, so presumably they’re congested. Asian nodes are often pretty slow as well, which IIUC is likely because for whatever reason bandwidth is particularly expensive at most cloud providers in Asia.
If you’re interested in a dVPN service, here’s my brief review of NymVPN:
Speed, connection issues, and all the basics vary. The “fast” mode with 2 nodes can offer decent download speeds, though this depends on many factors. Mixnet, which uses 5 nodes, is slower than a turtle.
I don’t feel secure when it comes to sensitive, critical information. For reading basic articles, watching videos, and things like that in “fast” mode, it’s worth it.
NymVPN has potential, but it has major drawbacks and weaknesses that are far worse than those of a higher-quality VPN.
Do I recommend it or advise using it?: NO. Reasons?:
They do not protect genuine users who have done nothing wrong from adversaries who have infiltrated their nodes and act according to their own needs. Their terms of service are weak. The company washes its hands of the matter—it’s your fault, not the company’s.
Nodes can be anyone, from a novice to state intelligence agents. This includes, but is not limited to: hackers of varying skill levels, the mafia, novices using AI to try to bypass the security of their own personal node in real time, APTs (advanced persistent threats)—which are difficult to prevent—and so on.
An exit node can manipulate your traffic if it knows how the Mixnet software works internally. Both government agencies and other groups can do this, as they have far more dedicated resources and time.
It is not suitable for emergency situations. Activists, researchers, etc., should not rely on these services for high-risk use.
Its “guaranteed” technical “theory” lacks fundamental evidence based on real-world facts. Parpers are not real evidence against state actors because state actors are dynamic; they adapt, play by their own rules—not Nym’s—and always have more dedicated resources.
Who is this service for? It’s for people looking for the basics or wanting to explore something “new,” like Mixnet, for example. For sensitive data, it’s not recommended—not even with Mixnet. Don’t be fooled by their advertising.
If your data is sensitive and you want a provider that doesn’t exaggerate or lie within the limits of actual technical capabilities, that’s IVPN; it’s the only provider I truly recommend—there’s no other.