And nobody is forced to upgrade to v4.
You know, I like DVT a lot, I use it not often but when using it I like it.
But I use it private. It is not the only subscription that wants some bucks from my wallet.
I have to calculate if it still worth.
Of course, the developers did the same. Calculated. And offered this "more flexible and modern license model.”
It would have been nice to see a new “more flexible and modern license model” which is placed little bit lower in price than this one.
Nobody likes price increases, but I’m not sure what the cheaper, better alternative is.
Evernote is $130 per year, and still doesn’t support any kind of Markdown.
Obsidian starts at $48 to sync one vault (database) of up to 1GB. I’d be looking at the $96 per year tier for my current usage.
Bear is only $36 a year, but it only syncs via iCloud and the sync isn’t encrypted, which makes it a non-starter for me.
As I recall there used to be three tiers of DEVONthink, with an entry level version for casual users. Maybe that would make sense again?
There have been three tiers for longer than I’ve been with the company. Nothing has changed there. There is Standard, Pro and Server.
If you only need it for personal purposes then probably the standard version would suit you most? 50€ once you decide you need the new features doesn’t seem too much. DT is not a tool that was ever known to be buggy, hence the only reason to update would be new features (or a general OS upgrade that renders many things obsolete and affects other software as well – which means new licenses for other one-time-payment tools as well).
1 Like
Huh, I guess I forgot about Server, turns out I’m using Standard. I think what confused me is that I did have a Pro license nearly 20 years ago, but when version 3 came out I decided I didn’t need the extra features. (Looking at the comparison page the only two things I’ve wanted from Pro have been OCR and email archiving, so yeah.)
I’m seeing an upgrade price of $49. If that’s going to be the annual renewal price, I have zero problem with that. I use it literally every day, it’s rock solid, and search is always instantaneous.
Don’t forget to check out the comparison matrix…
One of the many wonderful things about the beta is that Pro and Server feature sets are unlocked for all users. If you’re on Standard, now is the time to archive those e-mails and OCR those PDFs; if you’re on Standard or Pro, you have to try the new enhanced Server, not least because you’ll probably never get another chance without paying the steep differential (which is absolutely worth it if you have a use for it, but normally the only opportunity to try it out first is during the original trial period before you license up, and most people don’t get that far). So far the improvements to Server are my favourite thing in v4 and worth the upgrade just in themselves.
3 Likes
I would partially agree only if AI credits were included within the “purchascription” price. That is apparently not the case. I actually like the “bring your own API key” approach, but that does not justify the higher price tag more than any other new feature does, since the AI is not run on servers owned or paid for by DEVONtech.
I do not think the new pricing model is “unfair” in the literal sense. Assuming the year-on-year update price is always $99 for the Pro edition, it’s on par with the “industry average” subscription price of $8~$10 per month. DT4 also has the advantage of being a purchascription, instead of a subscription.
I’m deeply disappointed by the way the DEVONtech team presented their switch to a purchascription model.
- “Modern” in the marketing sense is a filler word for when a company can’t find any other good thing to say.
- That “if you extend directly, you will receive a discount” is the most unsettling part of the change, partly because we have not yet seen what this “discount” actually means, and partly because it bothers the user with a dilemma that makes zero sense from a UX perspective.
I have no regret with my purchase of DT3 several years ago. That wonderful, reliable app has been a vital part of my work. I see I don’t have to upgrade to DT4 at all if I (1) don’t need all the AI stuff, and (2) can do without auto versioning. So I’ll stay on DT 3.9.9 until migration to other software.
3 Likes
Actually it’s more flexible for two reasons:
-
It’s the user’s decision when to extend a license (contrary to subscriptions) as the software remains fully functional
-
We are now able to ship new features whenever they’re ready, no more need to delay stuff for larger upgrades
4 Likes
More flexible?
Sorry, but I don’t agree at all:
- “user’s decision when to extend a license”: until now, this question didn’t arise. The user just had to opt for an upgrade, not for a (disguised) subscription. And if I did’nt upgrade, my version remained also functional.
- “we are able to ship…”. Yes “you” but what’s in it for the user? Nearly nothing…
2 Likes
This statement makes no logical sense. Who do you think the additions, improvements, and fixes are for? They’re certainly not for us. And this change allows us to deliver these things to you more frequently than before. That is also benefitting you, not us.
7 Likes
sorry, I misspoke (I’m not fluent in English).
I just want to say that personally I’m not a month away from receiving the new releases.
And I don’t see how your new licensing system changes anything…
1 Like
The first, simple AI support in DEVONthink would have been available in 2023 for example (and other features sooner too).
2 Likes
Well, here’s a real world example: DEVONthink To Go 4 will come with a slew of new features. Some of them exist since autumn 2023. But because we also have a one-time purchase, not a subscription, we keep them for the paid upgrade, of course. You could have had them already more than a year ago but with the traditional upgrade model you have to wait until the next big release.
Our license model on the Mac changes this. New feature is ready, you get it right away, not in three year’s time. And unless on iOS, it’s not a must, it’s a yearly “institutionalized” upgrade cycle instead of huge upgrades every few years. So there’s almost no difference except that we are now able to ship features when they are ready, and that the cycle is always one year, not three, not six.
13 Likes
My rational to this is guided by what I already do with Bookends. It has a similar license to DT4 (though for 2 yrs usage). I normally pay every 3 to 3,5 years. So, for awhile I’m stuck with a fine version that works well, just not with the last improvements/features. I could use Zotero which is open source; I tried it, but for me Bookends is much better and worth the semi-regular payment.
This is what I plan to do with DT4. I know it is unavoidable that I’ll be paying more in the long run than I did for DT2 or DT3. (I trust that I’ll also be getting more from it.) But I can decide how often I want to pay and make it less expensive for me than yearly payments.
That said, if DT was not part of my regular flow, I wouldn’t have paid for DT1 > DT2 > DT3. I love to use it and it keeps my huge amount of academic data organized, which helps me a lot as a researcher and as a professor. In fact, with DT4 I will probably be using it even more (I’m sure Ai will help with my workflow, as will many of the new features in DT4).
But that is me and my needs and how much I value DT4. If DT is not that important in your workflow, the price is probably too steep and not worth it.
Now — my wish is that the promo system that DT offers from time to time (Black Friday for instance) will extend to academic licenses (i.e. that the 25% off can be applied to it as well). That would for sure entice me to pay more regularly for the one year update system. Will you consider it? (I don’t think that is the case now.) @BLUEFROG
4 Likes
We could do this, but we have an extremely high percentage of academic users. This would lead to discounts beyond what we can offer. Unlike normal people, as academic you already benefit from discounts all year long and there’s also the Back to School discount.
4 Likes
That is incorrect. It was an example of a monthly cost that would be easily relatable to many people, also evidenced by the inclusion of other services people may use. And also examples of not hugely exorbitant monthly costs, e.g., a car payment or cell phone, in line with the amortization I was showing. There was no explicit or implicit connection to (1) a subscription or (2) “streaming services”. And the monthly cost doesn’t imply it either, if you read my comments. It is merely to examine the cost in light of how it breaks down per month and per machine.
And again, it is not a subscription.
- A subscription would be: It will be $8.33 per month to run DEVONthink 4 Pro. If you don’t pay us each month, you’re locked out of your data until you pay up.
- Our model is: Here’s a year of all the updates – additions, improvements, and fixes – for $100 for the Pro edition. At the end of the year, you can continue getting them for another year for $100. If not, feel free to keep running the version you paid for and have all the access to your data unfettered. If after awhile you want to extend the license, you can update for the same $100. Now you have another year of updates.
7 Likes
One more time, it’s not a subscription model. Your purchased version remains fully functional for as long as your hardware and OS supports it, with no further payments. You pay for updates when you feel the updates have become compelling enough to warrant it. This kind of model has been standard for years on a whole range of apps in the same general space as DT, and if you don’t need the new features (as I haven’t for years with e.g. Tinderbox or Aeon Timeline), then you stick with the version you have. I pay the annual fee for updates to a handful of apps (Wotja, Bookends) because I value the new features and want to support development, but there’s no need to feel pressure to keep up with every new feature release unless it’s really compelling for you.
I for one welcome the move to a more regular, responsive update cycle. This weekend has been completely exhausting and I’ll be glad not to see another major update like this. It’s been a struggle just keeping track of what’s been said where on the forum, let alone trying out all the new features, and I’m only a user; imagine what it’s been like for the team making all this happen. DT3 was perfect and I felt no need whatever for a new version, but DT4 has blown that out of the water and I paid for the upgrade in a heartbeat. But that doesn’t mean I have to keep paying annually ever after, nor should you.
10 Likes
That’s a very interesting and valid point with which I, for one, certainly concur.
Stephen
4 Likes
As a regular visitor to these forums, you sometimes come across topics that are completely cryptic. Here it is the opposite: quoting the “more flexible and modern licensing model” and adding just a question mark get to the heart of what immediately struck me as not right when I first read it in the newsletter announcing the DEVONthink 4 beta.
I sincerely hope you consider your advertising of DEVONthink’s new pricing model as being in beta too.
Let me explain myself:
“Modern”: “Modern”[1] does not equal “better” or even “good”. Not in general and especially not in 2025. Due to the actions of a certain major player the trading environment has shifted from a (at least to a certain extend) cooperative way to “I want it my way and I will force it on you because I can.”
So “modern” to me seems to be a very unwise choice of words for a company that has never treated their customers like that (and that does explicitly include the new pricing model!).
“More flexible”: More flexible than what and for whom?
No. 2: Of course that’s a good thing (see also below).
No. 1: I’m really sorry, Christian, but this strikes me as nonsensical. Why are you comparing DEVONthink’s new pricing model to a subscription? DEVONthink (the Mac version) never was subscription based. “The new DEVONthink pricing model is more flexible than a DEVONthink pricing model that never existed”?
And compared to the old pricing model the new one is not “more flexible” for the customers—in regards of flexibility for the customer it’s the same!
Pricing model up to DEVONthink 3: On day X the customer purchases a licence (or an upgrade if they have a licence for a previous version) of DT with the specific set of features on that day.
They will get updates with bug fixes, to keep DT up to date to work with the latest OS, and maybe including new features until day Y. After that they have the option to decide wether to use their last updated version with all its day Y features as long as they want or the hardware/the software allows. Or to upgrade—immediately or some time later for a reduced price over the full version.
Pricing model DEVONthink 4 and up: On day X the customer purchases a licence (or an upgrade/extension if they have a licence for a previous version) of DT with the specific set of features on that day.
They will get updates with bug fixes, to keep DT up to date to work with the latest OS and maybe including new features until day Y. After that they have the option to decide wether to use their last updated version with all its day Y features as long as they want or the hardware/the software allows. Or to extend—immediately or some time later for a reduced price over the full version.
See? Same model. Definitely no subscription (not even “in other wording”). But no “more flexible” either.
The difference does not lie in the model but in the period of time between day X (day of purchase) and day Y (end of updates). Before DT 4 day Y was the day when the you, the developers, believed you had piled up enough new features up to justify a new major version. Whenever that would be. From DT 4 on day Y is exactly one year after day X.
And now comes the difficult part, the figures:
DEVONtechnologies has spoilt us. Badly.
It took you about a decade to upgrade from DT 2 to "DT 3. I do remember that at some time in the second half of that decade I began feeling bad about getting updates that did not just include fixes to make DT work with the latest macOS—fixes we would be so happy to get for other software—but new features as well without having me to pay for.
Wether I felt bad more because I thought that you simply deserved to get properly paid for your amazing work or because of more egoistical reasons—when you went broke, DT wouldn’t be under development anymore, what would I do without it once it would not work anymore?—from my point of view does not really matter.
Because I very “unmodernly” believe in cooperation. When I want to use a software (especially one intended for long-term use like DT) there must be a financial arrangement that is healthy for both sides. If a software is too cheap, the developers close shop. If it is too expensive, I can’t afford it anymore. In either case it’s bye-bye.
So ten years without paying clearly was on the bad side. When I read in the recent discussions that DT 3 was released six years ago first I couldn’t believe it, but my 2019 bill proved me wrong. Six years without paying was also too long—I guess at least the vast majority of DT users will agree.
The practice of DEVONtechnologies has been to only release a new major version of a software when it’s ready. That is exemplary but at the same time business wise quite risky when it takes that long.
In my opinion you took a (the only?) reasonable way out of that dilemma by getting away from the for-pay major version/free minor version scheme. Now we get a feature when it is ready, not a pile of features when they are ready, now you (probably) get a more steady income. Generated by existing customers but also by new customers, hopefully attracted more strongly by the constantly upcoming new features.
But since we still are in the figures sections of my novella, I should (not as the first here, of course) also state the figures clearly:
The extension cost me, currency fluctuations aside, the same as the last upgrade did. Even if we agree that we should compare the new pricing neither to the excessively long ten years between DT 2 and DT 3, nor to the six years between DT 3 and DT 4, and even if we assume that we could agree on a “normal” or “healthy” or “reasonable” time of, say, three years between two major versions, compared to that the price still has tripled.
Three times the price to have a version that is up-to-date. To put it another way: The same price for a version that is up-to-date plus not a whole stack of new features but just a few (as many as can get implemented properly in one year as opposed to in many years). In all fairness, this should be called what it is: a hefty price increase.
Don’t get me wrong: I’m not saying that DEVONthink is not worth it or that I am not willing to pay it. And I am not against this pricing model in general (while I would like to get rid of subscriptions better sooner than later).
But this pricing model has a weakness, and that’s Apple. I am not sure that it will work to “flexibly” decide whether to extend or not based on new features as in “Christian took a long vacation after the release of DT 4 and it was well deserved, but after one year only pastel highlighting colours are new, so no” vs. “Finally plug-ins that allow DEVONthink to do the dishes and take out the trash, so yeah”.
Apple is known for breaking with old methods. These don’t even have to be big things, like the switch from 32 to 64 bits. For example, there seems to be a macOS team or maybe a very, very dedicated individual employee for messing up PDFkit every year. And the developers of software that uses PDFkit, like DEVONthink, have to cobble together a workaround until everything is fixed in version .4 or so.
Other changes on that level might be permanent and thus require an update. The short DT extension period of one year and macOS that is also released annually hold a high chance of getting forced to purchase extensions. In that regard the aforementioned Bookends, same pricing model, just with a 2-year licence, is a bit more (i. e. twice as much) on the safe side.
But we’ll see.
What I wanted to say in so many words[2]: There are compelling reasons for the new pricing model. You should emphasise these and refrain from terms like “modern” and “the more flexible” for the users.
11 Likes