Tailscale's Pricing v4
tailscale.comReposting my comment from Reddit: Honestly, I’m a bit torn. On the one hand, I’m personally thrilled that the free plan now includes 6 seats, which is perfect for my home setup and honestly more than I expected.
On the other hand, it likely means I’ll lose the ability to use Tailscale at work. We’re a small startup with 8 people in our Tailnet, though only about half are active in a given month. Right now we’re paying somewhere between $6 to $12 depending on usage, but if I’m reading the new pricing correctly, that jumps to $64 a month. That’s a pretty steep increase for us.
Do you think the cost of evaluating replacements, agreeing on one, and switching all people and processes to this new tool would cost this startup more or less than 600$ this year? After how many years would that work break even at 600$/year?
I swag this work to take:
1 engineer 8 hours to honestly evaluate alternatives, their licenses, the likelihood they will move towards a similar model in the next ~3 years, number of stars on GitHub, etc
2+ engineers 2 hours to agree on the path forward
1 engineer 2 hours to implement to new solution
8 engineers 1 hour to switch to the new solution
1 engineer 1 hour to finalize the removal of the old solution
I’ll call it 24 hours of work total, how much are you paying your engineers? Let’s say an easy 100$/hr. At this rate, it would cost $2400 and take 4 years to break even on doing this work (assuming the new solution has feature parity at the old 12$/month rate)
If I was tailscale doing this math, I would bet your startup is not gunna do this work to migrate away from tailscale right now because of this price increase.
This is exactly why I will never create a business that sells to developers. Some of the highest paid individuals on the planet cannot fathom paying for tools that make their life easy.
I get there are lots of tools that all add up, but even paying an extra $2,000/yr per dev isn't really that much for the time savings. But for some reason we're fickle, we think things should be cheaper than a daily coffee because "we can create this ourselves".
> Some of the highest paid individuals on the planet cannot fathom paying for tools that make their life easy.
I get where you're coming from, but you do realize that developers aren't the ones making the purchasing decisions, right?
Yeah, my work is in a very similar boat. We might need to drop our Tailscale usage because of this. It sucks.
Oh noes… I’m a single-user but I have just a little under 50 tagged devices and a lot of ACLs. I don’t think my tagged device count is getting any smaller since I like to play around in my homelab. Been paying the Personal Plus subscription since I’m getting a lot of value out of Tailscale. Hopefully I can purchase some upgrades to the Personal plan.
We do intend to allow Personal users to pay for additional tagged devices; the ability to self-serve additional device counts is on our near-term roadmap.
This is potentiallya bit of a downgrade if you are using the Personal or Standard plans for infrastructure:
- Reduction from 100 to 50 tagged devices on Personal (was 100 total devices before)/Standard (was 100 + 10 per user before)
- ~~Limit to 1000minutes/month (16.6 days) for ephemeral devices (simply counted towards the total device limit before), meaning you can't even have a single permanently connected ephemeral device anymore on Personal/Standard~~ This is incorrect, see comment below
- Limit of 3 ACL groups on the Personal plan (previously no limit), but also allows 10 ACL groups on the Standard plan (previously had no access to the feature)
Unfortunately, the pricing for extra devices and ephemeral devices is "contact sales".
Note that if an "Ephemeral Device" exists for more than 4 hours continuously, it consumes 0 minutes, and is billed as a tagged device. The docs are being upgraded to change the language around this.
50 tagged resources is the new limit, but there is no longer any limit on user-owned devices, previously, they were bundled together in that 100 limit.
>Note that if an "Ephemeral Device" exists for more than 4 hours continuously, it consumes 0 minutes, and is billed as a tagged device. The docs are being upgraded to change the language around this.
Thanks for explaining, that makes the limit very reasonable. I updated my comment
So.. an ephemeral device becomes a regular device after 4 hours? First 4 hours count toward minutely quota and then toward device quota? Is ephemeral a toggle when creating the key itself?
If it exists for more than 4 hours, 0 minutes are consumed - the 240 minutes are "refunded" to the pot.
Ephemeral keys can still be used to manage long-lived devices when you want them automatically cleaned up if they disappear - they'll just count as tagged nodes when they hit that 241 minute age.
does the "tagged resource" limit include services?
No, services are separate. Services are currently soft-limited to 10 on all plans.
counts huh. well, I'm at 14 (which were previously just tagged "devices"), I guess at some point I get to find out what the hard limit is :).
Hey there! Tailscalar here, I'm a member of our Product team. We're working on how we evolve the Services pricing model, we know it's not exactly right at the moment. We don't intend to suddenly start enforcing a hard limit on folks who have been early users of the Services feature.
no worries :). I plan on remaining on personal plus anyway until such time as you stop letting me give you money (making this the second plan you'll have kicked me off of :P), but there's not actually any difference between the (now) state of that and the free plan as far as I can tell. (also is the plural of Tailscalar Tailvector?)
I'm on personal plus ($5) now. Should I change anything? I'm the only user in my tailnet and have 4 devices. I think I would probably be better served downgrading to the new free tier?
It's admittedly tough to know what features Plus had vs the new matrix
It looks like it depends on how much you use ACLs. The new ACL cap seems strictly worse than before. Otherwise it seems the new free plan is mostly the same as the old Personal Plus plan for most users.
I think the new free tier is going to give you a lot more value, and you can save the $5 :D
I have seen some demos online such as WebVM [1] that use ephemeral devices to enable networking within the browser. It sounds like there is now a limit on the total length of ephemeral sessions on an account that last less than 240 minutes. While I'm not currently affected by this limit, it does seem rather arbitrary.
[1]: https://labs.leaningtech.com/blog/webvm-virtual-machine-with...
Hi, lead dev of WebVM here. Can you link to the specific change that affects ephemeral devices? I cannot find this reference in the article.
Keep in mind that we also plan to offer builtin networking in the near future, we are developing the infrastructure to do so as part of our newest product (In-browser sandboxes): https://browserpod.io
https://tailscale.com/pricing says "1,000 mins per month for ephemeral resources"
Thanks for sharing. I see how this could impact heavy user, but there is always the option of authenticating via a non-ephemeral auth key as a potential workaround. It's not integrated in the public WebVM UI, but it is supported by the underlying engine (CheerpX).
With all the talk of simplifying things in the blog post, I had high hopes that were dashed against the rocks when I looked at the continuing insane complexity of the pricing page. Honestly, it's like looking at Charlie's murder board in Always Sunny. The one feature I use in Tailscale, ssh with magic DNS, seems to have unlimited hosts for the Personal plan and 5 hosts for the next step up. I would love to give you guys money, you just won't let me.
Introducing billing for ephemeral nodes and tagged resources behind an opaque "Contact Sales" wall doesn't feel like generosity, it feels like a classic bait-and-switch.
We architected our infrastructure around Tailscale (under their "now legacy" Premium plan) under the reasonable assumption that these specific usage patterns wouldn't suddenly become cost centers. For context, we run on-prem Kubernetes with Flannel as CNI in host-gw mode, using Tailscale purely as the underlying transport. Because of this architecture, every Kubernetes node acts as a subnet router, which now neatly falls into their newly monetized "tagged resource" bucket.
Because of this pricing change, we're now looking at a one-year ticking clock. Our options are to either walk into an enterprise sales negotiation at a severe information asymmetry disadvantage to keep our current architecture, or rip out our networking layer entirely. I've already added an "Evaluate Netbird" to our team's backlog.
So it's deeply disappointing, but perhaps we should have seen it coming. I already perceive this as the standard lifecycle of a VC-backed HN darling: build immense goodwill with developer-friendly terms, embed yourself as a deep infrastructure dependency, and then aggressively squeeze the margins once the lock-in is established.
Hey there! Tailscalar here, I work on our Product team.
Sorry about the Contact Sales bit; it's temporary. Our full intention is to allow self-serve purchasing of these things, it's on our near-term roadmap.
In your current plan, all devices are grouped into a single category, and you are allotted 100 devices plus an additional 20 devices per user in the tailnet. We've typically been pretty generous on enforcing this limitation. In the new plan, we have instead offered unlimited devices owned by users, and 50 owned by tags.
I'd love to understand more about your use case and understand how we can make the transition easier for you (not a sales call). Feel free to reach out sam [at] tailscale.com if you'd like to chat!
Respectfully I think this is a mistake. You should not have released the pricing update without making this self-service.
I've evangelised using Tailscale at our company and the 100+(20 * user) limit was very comfortable for us at our level of scale. Cutting that to 50 + "contact sales for pricing" means that our use-case is no longer supported. We use Tailscale as the control plane for remote access to embedded devices we ship around the world - we use tagged devices far more than user devices. I think most tech-focussed companies will also have far more tagged devices than user devices. Anything that discourages us from tagging a server or a random dev machine and putting it on Tailscale will reduce our use of the service and ultimately reduces the "stickiness" that you are probably aiming for.
I also think capping Tailscale SSH to 5 devices on the standard plan kneecaps the value. It's probably in the top 3 most "magical" features of Tailscale. Who has only 5 servers?
I'm happy to pay more if that's what is needed but I resent the "contact sales for pricing". I waste so much of my time jumping through sales hoops with vendors and have a strong preference for vendors that will openly tell me on their website how much my use case will cost.
What‘s the intended limits for Tailscale Services on the free Personal plan?