Tailscale could be improved in different ways. I have not really found any problem with it so far. It pretty much solved the problems of other VPN clients that we have been trying. If Tailscale provided a way to integrate with native cloud providers, such as AWS or other cloud platforms, it could be a very nice thing. For example, I want to connect my EC2 instance from AWS over my private Tailscale network. It could be a nice feature so I do not have to stress about the AWS VPC and all that. I could probably just port it over to my Tailscale network. I do not think they have a feature like that, but that would be fine. I rate Tailscale an eight because I do not think it has any native cloud support with AWS tools such as EC2, GCP, and Azure. If they can find a way to do that, then it would be fine. One of the things that would actually make more sense is if Tailscale could find a way for us to use serverless compute over private networks. Private server functions or serverless functions probably would not be executed on non-EU regions. That is something I am thinking about.
I think if there was a streamlined mechanism to self-host the control plane, that would be a significant help in certain regulatory environments where customers feel unsure about having an external party control authentication onto the tailnet. Even with Tailnet Lock, there is still a concern that there are parts of the control plane that are opaque.
The only drawback I found is that the subnet routing option in Tailscale is too advanced and sometimes requires advanced networking concepts. A person needs to fully understand it to configure it properly.
I think Tailscale can be improved by enhancing the free side of things. It is really useful for people who are students or personal users who are not business users to keep going with the free version because it expands your user base. If you can get people to realize that this is actually a good product, they will be willing to pay for the paid version.
Freelance Web Developer and IT Support Specialist at a tech services company with 1-10 employees
Real User
Top 5
Mar 24, 2026
One thing I want Tailscale to improve is their user interface for Arch Linux, which is one of the devices I have that I installed Tailscale on, but they do not have the native package with full features, a capability I really want them to develop. Another enhancement would be to allow a graphical interface for more power users. Sometimes I want to run things on the website or user interface, so I wish to control advanced access control lists with a good user interface, and that would really be helpful.
I would say that in terms of Tailscale, if I have so many tailnets I need to connect to, there are some issues in the login process that need a little bit of attention from the Tailscale team. Sometimes, you would probably need to restart your entire system for it to connect. As a DevOps as a Service engineer, I have so many clients that are using Tailscale, and I would need to connect to maybe five different tailnets because they are different clients. Client A might have a Tailscale, Client B, Client C, and so on. When changing between tailnets, sometimes it hangs. Sometimes you might need to restart your entire system. This is a bit of a headache in that aspect. However, I believe if you are just using a particular tailnet, just one, then it is quite easy and there won't be any headache. For me, I think it can be improved in the aspect of having multiple tailnets to connect to. A good refresh on the Tailscale side and the backend side to refresh the connection anytime there is a new connection to be made would be helpful. Instead of needing to restart the system, it should be able to refresh itself. The connection side and connecting to multiple Tailscale instances can be problematic, and sometimes you have to restart your system when switching between them. The ACL sometimes is like another language on its own entirely. It is fine, but they need to make it in a YAML format instead of the current format because it is quite new and something you have to go and study. If they can make it like a YAML format, that would be better. Aside from the switching which I mentioned and the fact that you have to relearn their ACL, if the ACL could be in a YAML format instead of JSON format, that would be beneficial. I don't think there is much they can do about the switching of tailnets, but if they can have a YAML format of the ACL, that would be good. Every other thing is a ten out of ten. The connection-wise is easy to set up and easy to install. It is good to have things connected all together from on-premises, from so many environments, and even exit nodes as well. It is good overall. The pricing, I think Tailscale can be a little bit on the higher side. It is not for teams with just small users. If you want to set it up for small users and a small startup, I don't think you can afford it and might need to go to other open-source alternatives. It is good for teams that have maybe fifty plus users or one hundred users. In terms of pricing, I would say it is on the higher side, but it is worth it. The price is worth the functionality. As a user, I would say it is more on the higher side, but based on its functionality, it is worth the price.
Data Engineer at a consultancy with 11-50 employees
Real User
Top 20
Mar 24, 2026
I still do not have any issues in mind that need improvement, but if I think about a new feature or something to be improved in the future, I may share it with you.
Field Service Engineer at a tech services company with 51-200 employees
Real User
Top 10
Mar 24, 2026
I do not think there is anything that I wish would have been better because, honestly, for my use case, it has everything I need. I read that people complain about the maximum number of users that can use it under one account, but that is not my use case. For my use case, it has nothing more that I need. It has everything, and it is perfect for my use case.
IT Support Engineer at a security firm with 11-50 employees
Real User
Top 10
Mar 23, 2026
Tailscale could be improved by having a better way to troubleshoot. Sometimes our devices do go offline, but if we are able to have some sort of command where we can instantly turn off and turn back on the services at the IP, that would be great. I believe that covers the needed improvements; it is already a pretty smooth experience.
I do believe there are a couple of features and changes I would like to see with Tailscale. I initially got introduced to Tailscale in high school when I needed access to services running on my server, but due to deep packet inspection being in place, I could not use Tailscale. Tailscale's free plan effectively introduces people to the service, but I would love to see an anti-censorship VPN protocol implemented. When accessing my homelab, I usually have to resort to using VLESS host through 3X-UI, which is complicated. I would like Tailscale to provide more censorship-resistant options, such as Shadowsocks or VLESS, as fallback protocols. I would like more anti-censorship protocols, such as VLESS or Shadowsocks, so I can effectively use Tailscale in environments with censored internet access that block WireGuard through deep packet inspection. More anti-censorship VPN protocols are the main improvement I wish for. If Tailscale implemented these, I would use it for all my VPN needs and would likely use Tailscale entirely.
Application Security Engineer at a tech vendor with 501-1,000 employees
Real User
Top 5
Mar 22, 2026
Tailscale does a great job for free-tier users, and the 100 device limit is very fair. However, I think improvements could be made in the support for UI clients, especially on Linux machines. In Windows and macOS, the UI support is good, but Linux is more customizable and has a unique setup, and I feel like I never have a good UI to interact with when using Linux desktops, which is unfortunate.
I feel the speed of the control servers are a bit too slow. I feel that's the main bottleneck right now for Tailscale. For example, at my apartment, I can do 2-gig networking, but when I connect via the exit node, and I'm wired in somewhere else, and that place also has 2-gig networking, I'm at max getting half of my speed. The main bottleneck here is really the control servers and the throughput of data. I feel Linux needs its own UI client. I had to use a custom third-party one. That is a big thing as well. To make it a 10, I mean, have better support for Linux. That's probably the main thing, honestly. Fix the DNS things too. There's some issues where I should be able to use my own custom DNS easily and then when you move MagicDNS starts causing problems, I should be able to just have my own custom DNS that links directly into Tailscale and assign each thing its own specific hostname. I feel that doesn't work as well as expected, or maybe you do provide that, but then it's kind of obfuscated through weird documentation.
Regarding how Tailscale can be improved, I think for free users, there are some limited options. However, we have a paid policy, so we pay Tailscale every month. At some point, we may want to host our own coordination servers, which Tailscale does not have right now. However, it is just a general consideration, and I doubt that we will have that problem soon. Regarding the needed improvements for paid users, I think things are acceptable. The limits could be higher for free users, and that is all.
Tailscale offers secure, remote network access without complex VPN setups, using Zero Trust architecture and WireGuard. It suits enterprises and personal use by simplifying connections across platforms and improving workflows.Tailscale enhances remote access by providing seamless integration across devices. It supports connectivity for subnet routers and exit nodes and integrates with GitHub Actions to streamline workflows. Its Zero Trust and WireGuard-based security architecture ensures...
Tailscale could be improved in different ways. I have not really found any problem with it so far. It pretty much solved the problems of other VPN clients that we have been trying. If Tailscale provided a way to integrate with native cloud providers, such as AWS or other cloud platforms, it could be a very nice thing. For example, I want to connect my EC2 instance from AWS over my private Tailscale network. It could be a nice feature so I do not have to stress about the AWS VPC and all that. I could probably just port it over to my Tailscale network. I do not think they have a feature like that, but that would be fine. I rate Tailscale an eight because I do not think it has any native cloud support with AWS tools such as EC2, GCP, and Azure. If they can find a way to do that, then it would be fine. One of the things that would actually make more sense is if Tailscale could find a way for us to use serverless compute over private networks. Private server functions or serverless functions probably would not be executed on non-EU regions. That is something I am thinking about.
I think if there was a streamlined mechanism to self-host the control plane, that would be a significant help in certain regulatory environments where customers feel unsure about having an external party control authentication onto the tailnet. Even with Tailnet Lock, there is still a concern that there are parts of the control plane that are opaque.
The only drawback I found is that the subnet routing option in Tailscale is too advanced and sometimes requires advanced networking concepts. A person needs to fully understand it to configure it properly.
I think Tailscale can be improved by enhancing the free side of things. It is really useful for people who are students or personal users who are not business users to keep going with the free version because it expands your user base. If you can get people to realize that this is actually a good product, they will be willing to pay for the paid version.
One thing I want Tailscale to improve is their user interface for Arch Linux, which is one of the devices I have that I installed Tailscale on, but they do not have the native package with full features, a capability I really want them to develop. Another enhancement would be to allow a graphical interface for more power users. Sometimes I want to run things on the website or user interface, so I wish to control advanced access control lists with a good user interface, and that would really be helpful.
I would say that in terms of Tailscale, if I have so many tailnets I need to connect to, there are some issues in the login process that need a little bit of attention from the Tailscale team. Sometimes, you would probably need to restart your entire system for it to connect. As a DevOps as a Service engineer, I have so many clients that are using Tailscale, and I would need to connect to maybe five different tailnets because they are different clients. Client A might have a Tailscale, Client B, Client C, and so on. When changing between tailnets, sometimes it hangs. Sometimes you might need to restart your entire system. This is a bit of a headache in that aspect. However, I believe if you are just using a particular tailnet, just one, then it is quite easy and there won't be any headache. For me, I think it can be improved in the aspect of having multiple tailnets to connect to. A good refresh on the Tailscale side and the backend side to refresh the connection anytime there is a new connection to be made would be helpful. Instead of needing to restart the system, it should be able to refresh itself. The connection side and connecting to multiple Tailscale instances can be problematic, and sometimes you have to restart your system when switching between them. The ACL sometimes is like another language on its own entirely. It is fine, but they need to make it in a YAML format instead of the current format because it is quite new and something you have to go and study. If they can make it like a YAML format, that would be better. Aside from the switching which I mentioned and the fact that you have to relearn their ACL, if the ACL could be in a YAML format instead of JSON format, that would be beneficial. I don't think there is much they can do about the switching of tailnets, but if they can have a YAML format of the ACL, that would be good. Every other thing is a ten out of ten. The connection-wise is easy to set up and easy to install. It is good to have things connected all together from on-premises, from so many environments, and even exit nodes as well. It is good overall. The pricing, I think Tailscale can be a little bit on the higher side. It is not for teams with just small users. If you want to set it up for small users and a small startup, I don't think you can afford it and might need to go to other open-source alternatives. It is good for teams that have maybe fifty plus users or one hundred users. In terms of pricing, I would say it is on the higher side, but it is worth it. The price is worth the functionality. As a user, I would say it is more on the higher side, but based on its functionality, it is worth the price.
I still do not have any issues in mind that need improvement, but if I think about a new feature or something to be improved in the future, I may share it with you.
I do not think there is anything that I wish would have been better because, honestly, for my use case, it has everything I need. I read that people complain about the maximum number of users that can use it under one account, but that is not my use case. For my use case, it has nothing more that I need. It has everything, and it is perfect for my use case.
Tailscale could be improved by having a better way to troubleshoot. Sometimes our devices do go offline, but if we are able to have some sort of command where we can instantly turn off and turn back on the services at the IP, that would be great. I believe that covers the needed improvements; it is already a pretty smooth experience.
I do believe there are a couple of features and changes I would like to see with Tailscale. I initially got introduced to Tailscale in high school when I needed access to services running on my server, but due to deep packet inspection being in place, I could not use Tailscale. Tailscale's free plan effectively introduces people to the service, but I would love to see an anti-censorship VPN protocol implemented. When accessing my homelab, I usually have to resort to using VLESS host through 3X-UI, which is complicated. I would like Tailscale to provide more censorship-resistant options, such as Shadowsocks or VLESS, as fallback protocols. I would like more anti-censorship protocols, such as VLESS or Shadowsocks, so I can effectively use Tailscale in environments with censored internet access that block WireGuard through deep packet inspection. More anti-censorship VPN protocols are the main improvement I wish for. If Tailscale implemented these, I would use it for all my VPN needs and would likely use Tailscale entirely.
Tailscale does a great job for free-tier users, and the 100 device limit is very fair. However, I think improvements could be made in the support for UI clients, especially on Linux machines. In Windows and macOS, the UI support is good, but Linux is more customizable and has a unique setup, and I feel like I never have a good UI to interact with when using Linux desktops, which is unfortunate.
I feel the speed of the control servers are a bit too slow. I feel that's the main bottleneck right now for Tailscale. For example, at my apartment, I can do 2-gig networking, but when I connect via the exit node, and I'm wired in somewhere else, and that place also has 2-gig networking, I'm at max getting half of my speed. The main bottleneck here is really the control servers and the throughput of data. I feel Linux needs its own UI client. I had to use a custom third-party one. That is a big thing as well. To make it a 10, I mean, have better support for Linux. That's probably the main thing, honestly. Fix the DNS things too. There's some issues where I should be able to use my own custom DNS easily and then when you move MagicDNS starts causing problems, I should be able to just have my own custom DNS that links directly into Tailscale and assign each thing its own specific hostname. I feel that doesn't work as well as expected, or maybe you do provide that, but then it's kind of obfuscated through weird documentation.
Regarding how Tailscale can be improved, I think for free users, there are some limited options. However, we have a paid policy, so we pay Tailscale every month. At some point, we may want to host our own coordination servers, which Tailscale does not have right now. However, it is just a general consideration, and I doubt that we will have that problem soon. Regarding the needed improvements for paid users, I think things are acceptable. The limits could be higher for free users, and that is all.