Steam Deck Software in 2026: Checking In With The Developers Behind The Ecosystem

The Steam Deck has been around for a few years now. In handheld terms, that's long enough for the ecosystem to settle, communities form, projects to grow up, and for the people building software around a platform to figure out what works, what doesn't and where things might be heading next.

Despite a flood of 'competing' handheld PCs arriving in the years since launch, and even Android devices increasingly pushing into PC gaming territory, the Steam Deck still feels like the center of the handheld PC gaming world as we move through 2026. Most of the software people use on it every day didn't even exist when the Deck first launched. Projects have grown alongside it, evolving from small utilities into essential parts of the experience.

So, approaching the middle of 2026, I thought it might be interesting to check in with a handful of developers building software for the Steam Deck scene. Most are friends I have spoken to before here on the 'Review, and one new face as well, to see how things are going, how their projects are evolving, and what the state of Steam Deck software looks like today.

You'll hear from RetroDECK, Junk Store, Heroic Games Launcher, Decky Loader, Lutris and Unifideck. This is my mini State of Steam Deck Software (2026 edition)!

RetroDECK:

Lazorne and Xargon are repsenting the RetroDECK team, sharing their statements with me

How do you view the Steam Deck ecosystem in 2026?

Lazorne: Valve’s progress in Linux gaming has been remarkable. With Linux gamers now representing nearly 5% of the Steam user base, the ecosystem is thriving. This momentum has also inspired a range of spinoff projects, such as Bazzite and Nobara, while encouraging greater standardization efforts across multiple Linux distribution communities. These groups are increasingly able to share technology and collaborate through initiatives like the Open Gaming Collective.
It had been a long time coming since the first Steam Machines and Valve’s initial investment in Linux during the Windows Vista era.
The Steam Deck is more than just a handheld device, it represents the culmination of Valve’s decades-long commitment to Linux gaming.
Thus the ecosystem is great!

Has the community or software landscape changed in ways that surprised you?

Lazorne: Yes, I think the growth has been even greater than I initially expected. More developers have become passionate about supporting Linux and FOSS, driven in part by the poor reception of Windows 11 and a growing awareness of the software people run on their hardware, alongside the major improvements in game compatibility on Linux.
The Unity ”Fee Controversy” also pushed many game development studios to reconsider their tools, leading some to move toward Godot as a free and more ethical alternative. That shift has further strengthened the broader open-source ecosystem.
Linux and software freedom are no longer just niche topics discussed by computer enthusiasts. Increasingly, even non-technical users are beginning to appreciate the benefits they offer from greater control and transparency to privacy, customization and long-term control of their devices.

Is what you make still as relevant now as it was the day it was released?

Lazorne: One of the biggest reasons I bought the Steam Deck was Valve’s long-standing track record of supporting its hardware. I did not know whether it would become such a major success, but I remembered the Steam Link. Valve released the hardware back in 2015 and discontinued it in 2018, yet even eight years after its discontinuation, it is still receiving updates. The latest build, version 918, was released on February 25, 2026.

The device still works flawlessly, allowing me to stream games from my gaming PC to my old TV without any issues.
The Steam Deck has been a far greater success than the Steam Link ever was and Valve continues to add new features and improvements on a regular basis. Very few companies in the industry support their hardware with such a long-term strategy, and for me, that alone makes Valve’s devices stand out above the competition. A device may have more raw performance or better battery life, but if its firmware and software eventually become obsolete, the burden falls on the community to keep it alive.
What makes Valve different is that they embraced openness from the beginning, giving enthusiasts and tinkerers the freedom to modify, repair, and extend the lifespan of their devices instead of locking them down.

RetroDECK is something many Steam Deck users interact with every single day. Maybe not every owner, but certainly a large portion of the community. How is RetroDECK doing in 2026, and how has the project evolved alongside the growth of the Steam Deck itself?

Lazorne: RetroDECK has always had a clear vision: to create a unified retro gaming platform within a contained sandbox environment. Since the beginning, we set ambitious goals and milestones for what we believed was necessary to achieve that vision and we have continued to grow our user base year after year.

Our next major release will mark the completion of several of those long-term goals. Version 0.11.0, nicknamed Everything Explosion, will be the biggest update we have ever released. In many ways, it is just as technically significant as 0.10.0b, building further upon the subsandboxing systems we introduced in that release.
Over the years, we have effectively built one of the most technically complex Flatpak applications in existence. Along the way, we have had to invent entirely new methods to solve problems that arise from pushing Flatpak to its absolute limits.

At the same time, RetroDECK remains a small passion project maintained by fewer than a handful of people in their free time. Despite that, we are incredibly proud of what we have achieved. Personally, one of the most rewarding aspects has been connecting with so many amazing people in the FOSS community who share the same passion for software freedom, preservation and open technology.
People often say the FOSS scene is filled with drama... and sometimes it is. But it is also full of heart, collaboration and community.
I’m thankful that so many people are enjoying something we have created.

So, to everyone out there: thank you for making us feel the warmth of the FOSS and RetroDECK community, especially in what can often feel like a very bleak world.

Xargon: While I think that projects for gaming on Android are great, my phone is nowhere powerful enough to play modern Android games, and with PC games it struggles even more, but I personally use GameNative for less demanding titles, and, albeit I prefer playing on the Deck, I could test some games and I am very amazed they could actually run in a playable state.

I just hope that the gaming community won't move on Android as now "Google Android" is the market standard and I see Google trying to restrict the access to our devices more and more creating a toxic ecosystem that should have been FOSS.

Locked bootloaders, Google Play Services and Google Play Protect are an extremely good way to turn the users into digital slaves, I personally believe that private companies should not have this power over their customers...but this might be a good case for the European Parliament and maybe this is off topic here.

So back on topic: talking about GNU/Linux ecosystem, I was amazed on how is growing thanks to the FOSS community efforts, I recently discovered Amethyst Mod Manager. Imagine Vortex + ModOrganizer2 fused together and Linux native: I had a wonderful experience setting up a 550 modlist for Skyrim.

Another great project, is Faugus Launcher I daily use it both on my Deck and on the desktop: ever wanted to just double click on a windows exe file to run your game? That's pretty much Faugus Launcher.

Put this together with the RetroDECK's RetroENGINE and you could possibly play any game ever released in human history on a single Linux PC with just a double click.

"Aaaah! Everything Merlin?" (quote, Archimedes - The Sword in The Stone)

Well, not everything, but probably we're getting there. And talking about RetroDECK, with the new framework, formerly codenamed "Project Neo" I finally I can take a breath: earlier, every emulator added or simply updated was a real pain to manage and accustom together with the others, now I can both focus on the actual components (and general RetroDECK's) development instead of wasting ton of hours to fit everything in a too small box. Moreover, the time I don't consume in that former frustration is time I can use for family, relax, gaming or such, so a great win.

I take the chance to thanks again all the other RetroDECK developers, contributors and testers that made this possible. Thanks to that lately we released 8 minor versions (patch, updates, bug fixes) to the RetroDECK 0.10.0b, reaching 0.10.8b, and now we're working on the 0.10.9b as possibly the last minor patch before 0.11.0b, that is actually already in the works.

With love!

//RetroDECK Team

Junk Store:

Eben Bruyns is represnting Junk Store

We’re now around four years on from the release of the Steam Deck, and despite a huge wave of competing handhelds arriving in the years since (and even Android devices now being capable of running PC games) it still feels to me like the centre of the PC handheld scene.

How do you view the Steam Deck ecosystem in 2026?

The Steam ecosystem (beyond the Deck) is very exciting. Valve still hasn't counted to three, but I feel more confident that there will be a three (or maybe skip it and go straight to four, or maybe 10 for the ternary geeks). We still don't have a Steam Deck 2, but we have a Steam Controller (technically 2). The Steam Controller is actually responsible for the existence of Junk Store. Steam Input with DOS games was always the dream. The Steam Machine (technically 2 as well) is coming out soon™, and the Steam Frame (2 if you count the Index as 1). All signs are either pointing to a good future or everything is about to be cancelled (I hope not).

I think what is surprising me the most is how fast and at the same time slow things are moving. I've been working on Junk Store for over two years and there's still no first-party offering to help out or replace what I'm doing. I'm still pursuing partnerships with any of the storefronts who are interested in working with me. The market is clearly tough and nothing is as simple as we'd like to imagine it to be. I think Valve is firmly cemented in place for the foreseeable future. There's a whole lot of something that's also happening very fast but there's no dramatic shifts. This is both good for stability but far less exciting. It's a sign the market is starting to mature a bit.

Has the community or software landscape changed in ways that surprised you?

The community is growing, more and more people are switching to Linux, and honestly I never thought I'd see the day. I do think those of us who are Linux veterans need to be kind to newcomers and remember that "I figured it out" does not always translate to "easy". A little bit of patience and empathy will go a long way towards cultivating a larger happy Linux community. I recently saw a YouTuber mention this, with people telling him something is easy, then walking him through 25 random steps. We still have some work to do to make things truly simple and easy. I should know; I've spent countless hours reducing the complexity and "step count" in Junk Store.

The ‘software landscape changing’ is always an interesting discussion. I've been doing this for over 30 years now. I can honestly say even over that scale a lot has changed...but it's also still the same. What I mean by this is we constantly recycle old ideas with a new coat of paint. What is truly new, and it's too early to make a judgement call on, is the rise of A.I. coding.

There seem to be a few different camps and all have strong opinions. Personally I think we're still in the hype and honeymoon phase with this technology and we will only know which way it's going to break once the dust settles. One of my long-held beliefs is that there's no substitute for experience, I don't know how this will hold up in an A.I. world. Will A.I. change this? Honestly, I'm not sure.

If you'd asked me this question a year or more ago I would have laughed at it. The tech looks different today, but will the current version of the tech replace developers? Maybe some of them, but not all of them.

Will a future version of this tech replace all of us? It's very hard to say because the last 10% often takes 90% of the effort. Will the tech plateau? Will code turn to grey goo? These are all questions I honestly do not have a good answer for at this stage. There's no denying that from a pure technical perspective it's pretty cool, but we have to ask if the juice is worth the squeeze.

I've seen a new term, "Denial of Attention Attack". This is actually something I think we should take seriously, because if FOSS devs get overrun with A.I. pull requests, it will burn these devs out. The main concern I have is that reviewing A.I. code requires a much higher cognitive load than normal human-written code. Humans use a small subset of the language feature set and patterns; A.I. uses them all, so you have to be exceptionally well versed in reviewing this type of code. Personally, that would destroy the fun of it for me and turn it into unpaid work that I don't want to do. This is naturally my opinion. I reserve the right to change it, and I certainly reserve the right to be wrong.

Junk Store is something many Steam Deck users interact with every single day. Maybe not every owner, but certainly a large portion of the community. How is Junk Store doing in 2026, and how has the project evolved alongside the growth of the Steam Deck itself?

Personally I think it's very cool that I can create something on a small sailboat in New Zealand and it touches so many devices around the world. I certainly could not have imagined this when I was a much younger and more naive developer.

Junk Store is still doing very well. The FOSS version is still wildly popular, a lot more than people realise. I don't have exact numbers and honestly it's almost impossible to know, but I would like to share some numbers with you to give you an idea of the challenges we face. If you look at the Decky Store stats there have been over 300,000 cumulative downloads. If you look at the legendary flatpak downloads there are over 200,000 cumulative downloads. Our website sees over 250,000 unique visitors per month.

Granted these are just 3 data points and none of those point to how many people actually use Junk Store, but it does look like there's quite a few who use it.

Junk Store is still growing towards my grand vision, and you can see glimpses of it shining through in version 3. I still have a lot planned in the pipeline but I don't publish a roadmap because it creates pressure that is impossible to manage as a solo dev. My hope is that when I do release features I find cool, people enjoy them and find them useful.

I released a full-blown file manager in game mode. That might seem insignificant, until you try to go without it. The feature itself is boring, but the friction reduction is far more than I initially realised. These features are all coming into existence as I'm building towards the main goal of being feature complete. Stuff is missing and is required to get to that goal, so it grows unexpected and useful features along the way.

In this vein, Junk-Loader, which is responsible for the release of version 2, was born out of friction in my dev cycle against Decky Loader. While Decky Loader is a fantastic general-purpose loader, Junk Store is a very demanding piece of tech and requires something specialised. Once I stood on the shoulders of version 2, the next friction point was Valve updating the Steam UI frequently and breaking Junk Store. This led to the creation of a custom UI/Navigation stack. This decouples Junk Store from the Steam Client dev cycle dramatically and ensures it keeps running. I'm very proud of this piece of work, it was no small feat and it took thousands of hours of research to get here. We keep a track record of SteamOS and Steam Client updates and Junk Store Pro performs through these updates.

Is this completely safe from Valve breaking it? No, nothing ever can be. But it's a very big step towards that without Valve co-operating on it. The public tracker only shows what is publicly verifiable, even though it's been running internally for an additional 6 months with no issues. I'd rather you take the public record's word over the "trust me bro" version.

Is what you make still as relevant now as it was the day it was released?

Yes it's still relevant, and the tell is that none of my features have made it into Steam like some other plugin features did.

Junk Store Pro is still growing in adoption. It's nowhere near the FOSS version's uptake, but it's sufficient to support us. All the effort is going into Junk Store Pro. The architecture is much better and more manageable for a solo dev.

Junk Store Pro has also grown tremendously in its feature set. Version 3 required a file picker for some features, which rapidly grew into a full-blown network-enabled file manager. This has essentially allowed for adding anything to Steam as a non-Steam game by simply right-clicking (or left-triggering) on a file and adding it as a shortcut. This technically closes the last open loop, and now Junk Store supports everything from game mode.

Obviously we're not going to stop there. We have at least 4 or 5 major pieces of R&D that spawned post release, mostly because version 3 actually delivered all the R&D that was midflight with version 2. There are a lot of exciting things in the works, but it's still too early to talk about those.

I have asked Valve for a Steam Frame dev kit, but honestly I doubt they would send me one. They must know what I'll be doing with it, but on the off chance someone at Valve reads this: be a sport, let me unleash myself on this awesome tech! I can't wait to get my hands on the Frame and the Machine so I can make the new features work on those. I've got some ideas around the File Manager and the Frame, it could be interesting in many ways.

Heroic Games Launcher:

Linguin is representing the team, sharing their statement with me

The whole Steam Deck community has always been amazing, while I have fallen out of touch on the recent software advancements, I love to see new gaming related projects in the space. Even though some are strongly built with AI, it's amazing to see people building their own utilities or even full clients based on the work we laid out with tools that power Heroic, like our contributions to legendary, heroic-gogdl and nile.

Heroic is doing great, we have released some features, we think will make the project even more appealing for handhelds like Console Mode inspired by Steam's Big Picture mode. I'm personally hard at work developing some long awaited store integration. I have no doubts it will be welcomed by community (assuming I succeed with it), all I can say now are these emojis as a little hint - ❎📦🎮 🎫

Decky Loader:

Chatting to AAGaming who is a developer from Decky Loader

Hard to believe the Deck's been going for four years now! I still remember the struggle to get my LCD model reservation in back in 2021. I've seen a lot of new handhelds enter the scene, but the deck still feels like the tried and true contender and is still what I'd recommend to people who just want their games on the go.

The software ecosystem for the Deck seems to be as active as ever. I haven't kept up much with projects outside of the Decky sphere as of recent, but I've seen a lot of innovation with third party tools such as Heroic, Junk Store, and of course Decky's many plugins (the submissions are getting a little overwhelming...)

As for Decky itself, while there hasn't been much new development as of late, the project is definitely still alive, and we try to still keep compatibility with new Steam betas generally within a day from release. There's still a few more features I want to finish up (left sidebar menu API, potential desktop support, VR 👀), but also I think the project is in a pretty good state so for now I plan to focus on stability improvements and better error recovery (Hi Reddit, I see you!) and overall installation management options. I think we've done a pretty good job keeping up with the Deck and it's community as a whole, and even four years in I'm still excited to see all the new and exciting plugin ideas that people are still inventing. Here's to the future!

Lutris:

We’re now around four years on from the release of the Steam Deck, and despite a huge wave of competing handhelds arriving in the years since (and even Android devices now being capable of running PC games) it still feels to me like the centre of the PC handheld scene. How do you view the Steam Deck ecosystem in 2026?

The Steam Deck is still going strong in 2026, it has imposed itself as the standard in handheld PC for many reasons but the most important ones are: its price and the fact it runs SteamOS and not Windows. It is now a fairly common device to see in the wild whereas the other handhelds still seem niche.

Has the community or software landscape changed in ways that surprised you?

The Steam Deck has become the baseline for many game publishers, even big ones. Some games like Cyberpunk have a Steam Deck setting preset. Big names like Capcom or Sony make Steam Deck compatibility one of their selling points. It’s a big enough part of the market share that Steam Deck compatibility and by extension Linux compatibility is now an important factor for many game studios. Game developers need to make sure their game is optimized for current gen hardware; Playstation 5, Switch 2, Xbox are all obvious and traditional targets but the Steam Deck is the closest we get to a baseline on the PC market. This is beneficial to everyone and not just Steam Deck owners since even those with more modest setups will have a chance to enjoy new releases.

Is what you make still as relevant now as it was the day it was released?

4 years after its release, it still holds up really well against the competition. A lot of the newer handhelds will have a stronger CPU and GPU but also a higher resolution screen, and in a lot of cases 16GB of shared memory, like the Steam Deck. This means that many games will run worse on more powerful hardware since they have to render a lot more pixels. In terms of dots per inch, the resolution of the Steam Deck is already pretty good. Increasing the screen resolution for a small device is not always a good thing, it is costly in terms of battery life and frame rate. It’s the same thing for laptops and phones, constructors want to put the highest resolution possible because its a strong selling point but you end up having to run your device at half resolution to preserve battery life.

Lutris is something many Steam Deck users interact with every single day. Maybe not every owner, but certainly a large portion of the community. How is Lutris doing in 2026, and how has the project evolved alongside the growth of the Steam Deck itself?

The Steam Deck, along with Proton and of course Microsoft continuously shooting
themselves in the foot are all factors that have played a role in the growth of Linux gaming. All gaming related projects like Bazzite, Heroic or Lutris have benefited a lot from this surge of users. This has changed the landscape quite a bit, we now have a a lot of users who are new to Linux and are less familiar with the tools and ecosystem we take for granted. These newcomers can provide very valuable insight to find points that can be confusing. At the same time, more experienced Linux users will tend to continue developing the project the way it’s been done in the past. I want to spend some time refining the user experience on Lutris so it is simpler to use while not removing features.

Unifideck:

Mubaraknumann is representing Unifideck, sharing their statement with me

How do you view the Steam Deck ecosystem in 2026? Has the community or software landscape changed in ways that surprised you? Is what you make still as relevant now as it was the day it was released?

Slightly surprising, but also not, considering the global software landscape, but vibe coding. I see a spike in new tools and plugins recently and in my opinion, its mostly because of the gap between ideas and implementation that AI is filling. Given the age of the platform and the operating system it was built on, there are a ton of pains we deal with every day, ranging from slightly annoying to QOL to incompatibility and almost all of us have some sort of idea or solution to a particular problem that would (hypothetically) resolve it. But not every one of us has the technical expertise and experience to even try and see if it would work.

I am one of these people. I work in Software Implementation but have absolutely no coding experience. Without the tools we have available in <current year>, I would not be able to bring my ideal vision of a SteamDeck to reality. I know most gamers and the Decky Loader team specifically are very much against the use of AI but I dare anyone to use our plugin and call it 'AI Slop'. I think as long as the end product is useful, well thought out and properly engineered, we should treat it as any other piece of software.

Is what I make still relevant as it was the day it was released? Not relevant enough yet! We support 5 third party stores right now and working on increasing the number, so we're getting there and will continue to listen to community feedback. And I expect it to be doubly useful once the Steam Machine hits the market, so excited for that.

Are you against the Decky team's rules regarding A.I. code generation?

Well,

1. I have nothing against the Decky Loader team gatekeeping and having a rigorous set of requirements and standards. I would actually prefer they make it stricter to make sure less and less slop goes through
2. They already restrict kernel level code and you can’t unlock the filesystem unless you have admin access anyway, I don’t think the distinction matters there

The Decky Loader team have a blanket condition for submissions - No AI generated code allowed. But in this day and age, you would be hard pressed to find any code that was completely untouched by AI. Heck you won’t even find any IDEs that don’t have it enabled by default, so I think it’s a matter of getting with the times. And I don’t buy into their ‘stolen code’ argument either. How is it different from when developers used to go on to Stack Overflow and manually copy solutions from the platform, while losing a bit of their own dignity for even asking the question.

Unifideck has become something many Steam Deck users interact with every single day. Maybe not every owner, but certainly a large portion of the community. How is Unifideck doing in 2026, and how has the project evolved alongside the growth of the Steam Deck itself?

Yes, I was really surprised that so many people liked it and was really happy that I was able to make a positive impact for all the gamers out there. Your question just reminded me that its been 5 months since I released the initial version and its amazing to see how it has grown to where it is now, although still a bit rough around the edges. This has been in part to amazing contributions from new and veteran contributors alike, all of whom want to see this plugin improve. We are also working on a complete overhaul of the project architecture that will help future contributors navigate our codebase easily and also for us to integrate new stores and features much more efficiently. So expect rapid new improvements and further stability as soon as we have that in place.

All my thanks to everyone who put up with me nagging them to share their replies with me for this article!