the way humans are mostly using LLMs definitely still sucks.
that’s not where things have ended up currently. Claude Code and Codex are a bit different to “vibe coding”.
the open source projects are out there. and there’s more coming.
but yes, see you in a year. 
I don’t have sales numbers in front of me, but I feel like it’s probably been shrinking since 2022 or 2023. Same as a lot of other indoor hobbies, they had a boom during covid and now that people can go outside again, the average person is doing that and not spending money on new plugins to stay inside and work on music.
I can tell you that, annecdotally, I did NGNY last year while I was finishing my MS. I finished that in August and instead of going back to video games and music production (which is what I did in my spare time before school) I got into Jiu Jitsu and going out pretty much every weekend. Now I spend a combined 5-ish hours per week on music and videogames.
So it doesn’t really matter how good or bad software is, or how much it costs, I’m just not in the market for much of anything right now. I’ve picked up a few tools that help me work faster (think stuff like Musichack plugins) with my limited time, but I’m not looking at ProQ4 or C3 because I just don’t work on music enough to justify spending that kind of money on software anymore. I also already have a pretty deep library from when I was more active and I still think a lot of my existing tools are really good and I already know how to use them.
I suspect there are a lot of people in a similar situation to me. You get older and you realize that this is fun, but it’s maybe not the only thing you want to do with your life and that you have enough to do the things you do want to accomplish in this space. I make music that I enjoy and that impresses my girlfriend. That’s good enough for me, and my tools are more than good enough to support that.
building plugins right now and this thread hits close to home. i think White_Noise has it right - it’s not that people stopped caring about tools, it’s that the bar for “why would i buy this” got much higher after the covid glut.
what seems to be working is very specific functional things - tools that solve one problem you actually have, not another synth that does everything. the “do i need this” filter is ruthless now in a way it wasn’t in 2020 when people were just buying whatever looked interesting.
the market for “yet another lush reverb” or “yet another analog-modeled synth” probably is shrinking. the market for something that clearly addresses a specific workflow pain point seems to be doing ok. at least that’s the bet i’m making.
Oh it definitely is the case I think for most people, even those who do this for a living. I don’t really spend much on new tools anymore. At a certain point you fill up your toolbox and any new additions have to stand out enough/fill a need/purpose that the tools you already have can’t fulfill in order make the investment worth it.
The good news is it’s 2026! Nothing to worry about!
To flip this thread on it’s head, what WOULD I buy in 2026?
Short background, I work in primarily electronic music. Hybrid of software and hardware that heavily favors software in my actual finished product (I write and record my hardware for scratch ideas). I used to master semi-professionally (by which I mean I was never in danger of quitting my day job to do mastering full-time, but it was a decent side hustle). I’ve been producing for 10+ years, so I have some classic tools knocking around like Diva and Trash 2 (the good one!) and I feel my mastering suite is pretty up-to-date as of about 2 years ago.
I already mentioned Musichack. I picked up Fuel on black friday rent-to-own for like $3 per month. That’s a rounding error in my monthly budget for a solid option that I can just throw on every channel and get a mix into the ballpark. Sweet EQ is also kinda tempting, but I already own Tone Projects Michelangelo and Ozone 11 adv, so I have a couple different master grade color EQ options already.
Toneboosters stuff is great for getting my ProC3 fix at like 1/8 the price. I might pick up their multiband comp if I get more into mixing again, but then again I have Ozone’s multiband as an individual plugin.
If I were going to get into mastering more often again (my DMs are always open, feel free to reach out!) I would want to update my Ozone specifically for their AI unlimit function. I have a version of that which I can use online, but improving that workflow would be worth money to me if I had to do it more often. I’d also consider Limitless. I have a handful of limiters already and use Elevate most often. I don’t have another multiband limiter to choose from and Elevate isn’t always the right tool for the job, so Limitless would be my Elevate alternative. I have few HP/LP mastering filters from Airwindows, but TDR has their filter series now and the interfaces are nicer, and they have some good additional tools/options that my Airwindows plugs don’t have.
But again, at my current output of mastering 2-4 albums a year for friends, that’s not going to justify these upgrades.
There’s one or two plugins that don’t currently exist that would be insta-buys for me. But, I’ve talked with a few developers about them and they aren’t easy to do for one reason or another and end up very CPU intensive. Like, we’re talking these need to be dedicated rackmount units CPU intensive, because they’d eat a computer alive.
I beta test for Slate so I get a lot of free stuff too. Most of it does not end up in my workflow, it gets tested, confirmed working, and then gets filed away and not seen again or has one track where I mess around with it. Out of 15 things I’ve tested the last year, I think 10 or 11 were literally an existing plugin release in a new format/container, and checking UI more than anything. That said, I think this fits into the “addressing specific pain points” within Slate’s ecosystem because our complaint as testers for years has been that we have to load Virtual Mix Rack all the time and it takes forever to load, especially with multiple instances in a project. That sucks when you want to use one module, so Slate is releasing the VMR modules as standalone plugins now. Actually kinda smart, if you are already in the Slate ecosystem. I doubt that’s going to get a lot of new customers in the door, but it at least addresses a pain point for current customers.
And I think big companies are going to go more and more that way. NI doesn’t need 10000 new customers as much as it needs to keep the existing ones on its platform. Same for Waves, UA (I’m convinced this is the reason Native plugins exist), Arturia is getting to this point as well. Providing the massive existing libraries of plugins and sounds in the formats/sets that people actually want to install and use on their systems.
Kinda like the game space, indie is going to be where the real innovation is (think TDR, Tone Projects, Musichack, Kern [hopefully]) and they are going to have to be really creative to cut through the noise and get noticed. I’m excited for what that means - roguelites are a genre of game that hardly existed 10 years ago, and now 3 of the last 5 years my favorite new game has been a roguelite. So I’m actually super excited and optimistic for the world of plugins, even if the majority aren’t going to be something I rush out and buy.
This should be printed on posters and tee-shirts everywhere.
Seriously, if you make it, I’d buy it.
I wonder if the ebb is actually due to the pandemic glut. Basically a market correction. The tide came way in, now it has to go way out to reach equilibrium.
That is, I think people were stuck at home watching YouTube and buying stuff out of boredom, and probably bought stuff they didn’t actually need or want, and it takes a lot longer to sell off that stuff or recoup the expense than it does to make the initial purchase.
Also, I can’t help but notice that everyone I know had a lot more spending money in 2019-2020 or so. For instance, back then a slice of pizza cost a lot less. Also, I could still order a pizza without having to pay some random tech company more than the actual price of the pizza just for the ability to read the pizza place’s menu.
I’m not sure what the current state of a market is, but there is one huge thing to improve. I dunno, could it be just one open authorization standard inside CLAP or VST4? With public & private keys or something. “As a customer”, I would prefer to stick to one provider of plugins or software, because it’s just unmanageable.
p.s. no critics towards this developer, just an example
Mainly resentment over some mainstream DAWs incorporating gen AI may be driving it. Plus subscriptions and further enshittification. Also desires for people to be away from the screen in a world dominated by screens.
Not certain causes but strong drivers for it.
Personal experience - I have one app that’s clearly run by clankers and they keep pushing updates with new features that break other things. At the point of cancelling on them because of this, even though I like the fundamentals of the app if they’d just slow down and take care. And they also keep pushing “AI” features, which should make sense seeing as it’s a knowledge management thing, except it’s not reliable enough and doesn’t really add any value so I don’t use them.
So yeah, don’t think the fundamentals of the technology have improved at all, you’re basically rolling the dice every time - that’s pretty much what the technology still is at base level - and for coding that means the more you use it the more it becomes a certainty that something will break in a way nobody fully understands. Or on the other hand if you’re expert enough to fully review and understand everything, is there really that much of a value add - the way people have put it to me is “makes the easy stuff slightly easier, but makes the hard stuff way harder.” Related:
https://www.ft.com/content/7cab4ec7-4712-4137-b602-119a44f771de
Another good point I’ve seen made is that the current vibe coding hype is a lot about venture capital that’s heavily exposed to “AI” putting money into startups who in turn buy tokens to prop up hilariously unprofitable “AI” companies that mostly people use as a gimmick but won’t pay for, to make it look like they’re actually viable as going concerns.
So yeah, still on track for the bubble bursting icy mid-2027 as far as I can see, at least all things being equal. The main reason it won’t is either a tech breakthrough - like I say, no sign as yet - or that AI as a political project continues to be propped up. Like, it must be obvious that a lot of this is about consolidating power in the hands of the billionaire class, it’s pretty much their dream to rule through an army of robot slaves. A lot of the crazy investment and hype around a fairly lacklustre technology is only explicable on this basis imho. But whether they succeed depends on a lot of factors, including how quickly it happens.
I also think the PE bubble is about top burst, its often been one big pyramid scheme as new companies cover up the debts of bad companies (so they have to keep buying…)- A lot of software companies are caught up in PE money (notably NI et al, but who knows how many others’
‘’*Private equity (PE) faces significant challenges in 2026, including a massive backlog of over 31,000 unsold companies ($3.7 trillion value), slowing fundraising, and high debt loads from leveraged buyouts causing increased bankruptcy risks. Critics cite aggressive cost-cutting, asset stripping, and short-term focus on profit extraction over long-term growth…(source FT)‘’
I’ve been waiting for an Audiomulch update since 2010.
Still using it though…
