I never used AppGameKit before Version 2., so can't speak about the stability from then to now... but haven't had any major issues in terms of that.
With that said... there really isn't much difference between Version 2 and Studio.
And as Studio is what I more typically use., I will speak from that perspective ... but keep in mind that as they're essentially the same, consider it to also be true for "Classic" (Version 2).
Now I've been using AppGameKit since ~2017 (and became active in the community again ~2020)
During that time, the number of new Features / Functionality that I've seen added can be counted on one hand; unless we're talking about Mobile Exclusive Functionality to do with Monetisation.
Updates have in the past 12 - 18 months slowed to such a crawl that there have been (legitimate) questions raised about if The Game Creators have begun to Sunset AppGameKit.
And this is a problem, even *IF* AppGameKit is in Maintenance Mode; because Apple (iOS) and Google (Android) release regular updates to their SDK... we have found a few times support to output to these platforms has broken after the Platform Vendors release an update; then we have to wait Months (rather than Weeks or Days) for an AppGameKit Update to ship that corrects this issue.
It could be that said updates aren't easy to work around what is broken, but the lack of almost any communication definitely remains deeply concerning; especially for those who are using AppGameKit to produce Retail Products.
And it isn't that TGC are incapable of regular updates... as Game Guru Max (their current focus project) is seeing regular YouTube Update Videos, News, etc.
We have also learned in the past few months something many of us have suspected for a while., which is the head of the company (Lee Bamber) essentially believes that no one
wants to Program/Code anymore in favour of Visual Middleware.
I think this is likely based upon Sales Figures for AppGameKit., and I recall him saying something similar during the end of Dark BASIC Professional' Support and their switch in focus to FPS Creator and the original Game Guru.
The thing is the issue isn't that people have "Given up on Programming" in favour of Visual Development... rather the issue is with how TGC themselves handle Product Releases, Updates and Support., especially in the face of Competition that are either better Value or more Feature Rich / Complete.
Version 2., should've been a Retail Replacement *not* a Free Upgrade to Version 1... with a clear timeline of End-of-Life / Long-Term Support for Version 1.
It also would've been a great opportunity to increase the "Minimum" Bar from OpenGL ES 1.1 to OpenGL ES 3.2, as well as place various long-requested features in the "Update Timeline"
As a keynote; Version 2 launched in 2014., 3 years after Version 1.
Version 3 (Studio), was kind of released in this way... which was a good idea; the problem is that all but the Vulkan Renderer improvements made to Studio have also been implemented in Version 2 and said changes weren't *AS* Big; meaning they were basically taking away ANY reason for anyone to upgrade... couple this with the fact that the Retail Price was essentially doubled., but again with very little reasons for people to _actually_ upgrade.
I mean again they had the change here to upgrade the "Minimum" Bar from OpenGL ES 1.1 to OpenGL (ES) 4.3 / Vulkan 1.x.
I'd also argue that Version 2 / Version 3 should've expanded support from Desktop (Linux/MacOS/Windows) and Mobile (Android/iOS) to Console (Xbox/PlayStation/Switch)
In fact I'd argue support for the Switch would've been a MAJOR selling point and seen a MASSIVE increase in Sales; especially as a major barrier to Switch Development (unlike Xbox and PlayStation) is the need to purchase a Developer Console in order to Develop for the Platform... which isn't massively expensive but it is a bit long winded and can be difficult.
AppGameKit having a Player that can be remotely (via a Network) Broadcast to for Development purposes., which itself can simply be pushed onto the Nintendo Store; well that would open up the ability for ANYONE to then Test and Develop on their Retail Unit; before then jumping through the hoops to get a Developer License and Kit... and such is MUCH easier if you have a Working Prototype of WHAT you want to put on there.
I mean I made this same argument with Dark BASIC Professional back in 2007., when Microsoft was first opening up the Xbox 360 with XNA; and consider how popular that became with some great games being released in 2008 - 2012 (when it was shutdown).
Unity 3D did this very move in 2008., today they're not only one of the key "Go To" Middleware for Independant / Small / Budget Developers but have become an Official Partner of Microsoft Game Studios and integrated into Visual Studio.
Instead TGC started making partnerships with NVIDIA, Aegis and Alienware.
I mean the other thing to consider is that they integrated Virtual Reality (VR/AR/XR) into AppGameKit., they've made some Visual VR Editors using the AppGameKit Middleware; but we never saw VR as a Native Feature of AppGameKit (AGK VR is great., but it's a 3rd Party Expansion) and what's more is they were doing this when PlayStation VR was being announced and launched... along with most of the major VR Players at the time.
Had VR been added as a "Root" Feature in Version 3 (Studio)., had they pushed partnerships with the key VR Vendors offering support for Oculus, WMR, SteamVR and PlayStation VR... again while I doubt we'd actually have seen much in the way of titles released with it; we would've seen a noticable increase in adoption for Studio.
---- • ----
Look I know this was a long rant., especially when the question was basically a "Should I get this?"
But my answer is... if you're looking for something that you can use to create software that you plan to Retail., then I'd say... No.
TGC from what has been said and their actions., seems to have AppGameKit on the backburner.
Support for the Mobile Platforms is for the moment working, but next week an update could release for one of the platforms and break it again for months; as TGC seem to have little desire to maintain it; and that is ALWAYS going to be the problem with a Mobile Middleware, you either are Actively Supporting it *or* within 12 Months it'll become unsupported on Active Platforms.
It'll still be perfectly fine for Desktop Development, but then so is Dark BASIC Professional; and that's been unsupported for over a decade and is free.
I'd say if you're looking for something that is going to be supported., Unity or GoDot are better options.
They are more feature rich, they have a larger community in terms of support and tutorials... and they're in constant active development supported by teams who care and believe in the products.
You're not going to get that with AppGameKit.
As much as I wish that wasn't true... that's just the reality at present.
While we can point out the Small Development Team as being the issue here., I'd point to DBC/DBP/AGK during their Peaks; when TGC (or more specifically Lee) cares about a project., it's well supported and updated... when he moves on, well a product is left in limbo until they eventually decided to officially sunset it.
If you just want something to mess about with, then sure AppGameKit is worth it... there really isn't anything else like it on the market.
Arguably that's the biggest shame here too., because it's that uniqueness of a BASIC like Language with a reasonable powerful System Feature API that makes it an excellent Rapid Application Development, Independant and/or Beginner Development Platform.
It's a major gap in the market that TGC essentially have a monopoly on, and are failing to capitalise on.