GTA Trilogy Remaster Studio Grove Street Bets Its Future on BeastLink, a Multiplayer Kaiju Game With 250K Destructible Objects

Grove Street Games made its name porting some of the biggest titles in gaming history, from the GTA III era to the fairly recent Grand Theft Auto: The Trilogy – The Definitive Edition, but its latest project marks a sharp departure from that legacy. BeastLink is the studio's first original IP in over a decade, and it's swinging for the fences with a multiplayer kaiju action game built around one core idea: absolute, city-leveling destruction.

Set in a world where humanity clings to survival in shattered strongholds while colossal Beasts rule the ruins, BeastLink tasks players with scavenging resources, battling rival factions, and eventually "linking" with dormant kaiju to wield their monstrous power firsthand.

The game's main showpiece is Grove Street's proprietary SuperDestruction system, a fully networked, physics-based tech that lets players reshape the map in real time, with maps housing over 250,000 destructible objects and tens of millions of individual pieces. Between piloting a tank through crumbling streets or embodying a towering Beast tearing through skyscrapers, BeastLink promises chaos on a scale few multiplayer games have attempted.

With Early Access on PC (Steam) on the horizon, we sat down with Thomas Williamson, owner and CEO of Grove Street Games, to talk about the ambitions behind BeastLink, the technical challenges of building destruction at this scale, and what it means for a studio best known for ports to finally step into the spotlight with something entirely its own. Keep scrolling for the full transcript of our conversation.

• Grove Street Games and Its Work in GTA Trilogy - The Definitive Edition
• The Core Concept of Beastlink
• The Tech Behind the "SuperDestruction" and Server Networking
• AI Tools and Steam Content Disclosure
• Unreal Engine 5, Blueprints, and the Nintendo Switch 2
• Campaign Mode, Release Window, and Pricing

Grove Street Games and Its Work in GTA Trilogy - The Definitive Edition

Welcome, Thomas. First of all, can you tell us more about your studio?

Thomas Williamson: We started out as War Drum Studios in 2007. I was 21 years old, starting a studio, and I'm not sure how, but we were able to convince a few publishers to help pay us to do ports for different titles. We did Ghostbusters for PS2 for Atari, and we did Great History Channel: Great Battles Medieval from PC to Xbox 360 for a publisher in the UK called Slytherin.

It just sort of grew from there. We got into the mobile space pretty quickly once the mobile scene became available. We were one of the first game developers on the App Store; we had a few launch titles in 2008 when the App Store launched, and then we did mobile for a good while. In the middle of doing all the mobile work, we did a lot of console and PC technical contracting, but it seems like the majority of the contracts we were getting were mobile-oriented.

We really started to pick up steam when we did the mobile port of Ark: Survival Evolved from PC to mobile in 2017, I think, is when we launched that product. That one kind of kicked off our desire to go bigger in terms of scope for our projects. So, the next project that we did was the Grand Theft Auto Trilogy – The Definitive Edition remaster for Rockstar. That was a three-year project that ended up shipping across PC, console, and then eventually mobile SKUs. In total, there were like 72 total builds for that product when we were building it, which was a big change in scope from just doing mobile games. Since then, we've done a few more things with the Wildcard guys—we did another mobile port for them and also a Nintendo Switch port.

But in the meantime, we've been working on this title, Beastlink, over the past few years. We were really excited the month before last when we were able to finally show it to the public. I think the biggest, loudest response we got from the community was that everybody sees the potential. They see that there's a really novel game concept here that no one's really been able to harness before, and we're excited to try to tackle it.

Most people know Grove Street Games for Grand Theft Auto Trilogy: Definitive Edition, and there's definitely been a bit of a mixed reception to that game. How do you feel about it?

Thomas Williamson: I would say that's probably what most of our name recognition comes from. But before that, we had lots of fans. We've shipped probably a total of about a dozen different titles that hit worldwide number one on mobile charts. The game that we self-published, Auralux, actually has millions and millions of fans: we have like 10 million downloads on that game. We also have a big user base from the Ark Mobile port as well, which got probably 40 to 50 million downloads and a huge contingent of existing fans. So, I would say yes, but it's not as big as you would think. It's not like 95% of our name recognition comes from that, especially.

Regarding GTA Trilogy: Definitive Edition, I agreed with most of the people's reactions. That's probably about the most I can speak personally for how we felt about what came out and what our intentions were. But unfortunately, I feel like we did not agree with how the game was released and the response to it from a development side, and I think that would have changed the narrative significantly.

However, at the end of the day, looking at the behind-the-scenes metrics on those games, there were a lot of people who were playing them and really enjoying them. There was also a lot of people who wanted more, and I don't blame them. We all view these games as these really, really important milestones in gaming history, so I'm grateful to have had the opportunity to work on them.

I'm also in agreement that I don't think there'll ever be a perfect redux of those that's not done by Rockstar proper. If their Rockstar North team took it on and did some crazy thing, that would be awesome. But the reality of what it was also kind of mixed with people's nostalgia. You don't see these remasters coming out very often and getting complete praise, especially with really complicated, big gaming milestone projects like that. It's a really hard thing to do and we were cognizant of that when we were working on it. We knew that there was no way to make everybody happy.

The Core Concept of Beastlink

How did you come up with the concept of Beastlink in the first place?

Thomas Williamson: I actually have not been asked this question very much. I will say that the game's concept has evolved over the past few years. We knew there weren't many good kaiju games. We've been watching the Monsterverse film series do extremely well at the box office and noticed the complete lack of good kaiju titles in this space.

There's a lot of single indie games trying to do stuff with kaiju, but there are not any higher-end titles. One of the biggest reasons is that doing multiplayer destruction at scale is really hard. It took us a while to make something that works well, and I think that particular problem is probably a barrier to entry.

Also, the fact that it's sort of an unknown space makes it hard for investors to want to come in and spend money on it, because they just don't know how well it's going to do. So we're kind of forging a path here. There's a lot of unknowns for us based on how this launch is going to go, but the whole concept was: hey, there's this giant hole in the market, we want a kaiju game like this. We want to make a kaiju experience game that really takes advantage of being a human and a kaiju at the same time, takes advantage of modern hardware, and gives people that epic experience. That's our goal.

How do you balance humans against kaijus, though? That's gotta be hard, right?

Thomas Williamson: The biggest, easiest thing to do is to become a beast yourself, right? If there's an available beast on the map for you to take over and you've built up enough serum—that's kind of what we call our score system, where you score points in order to build up enough saturation to link—you can become a beast and fight another beast. That's easily the best way to balance it.

However, with vehicular combat, as in a helicopter, you can also spend some of those points to go into a fixed-wing fighter jet and shoot rockets into the kaiju. So, there are different avenues available to you. But at the end of the day, we did purposely make it so that the beasts are mostly meant to be avoided. They have extremely high HP and high damage, and you can't really do much against them unless you're another beast.

That makes them an interesting deus ex machina for different game modes, because if the beast is there, you might have to make a different decision about where you're going to go next. And if you want to team up with all of your buddies and do an 8v1 thing to take him down with a bunch of helicopters and air strikes, you can do that. It really just depends on teamwork and it depends on the game mode.

What can you share about the max player counts and how the squads are split? Is it humans versus beasts?

Thomas Williamson: No. We found that the best way to do it is to try to keep things even. All the game modes are designed for PvP, at least where you have two teams, and each team has equal access to the beasts. Anytime we did anything asymmetrical, the balance became a lot more difficult to manage, which makes sense because they're so powerful.

But you can do that in PvE. That wasn't available in the beta, but in the early access release, we'll have some PvE experiences. You can bring your friends and do a sort of quasi-campaign where you're doing humans versus beasts or beasts versus humans.

32 players is the maximum for 1.0. For early access, we're probably going to cap it at 24 until we're confident in the servers. The last thing I want to do is release the game and have it not perform well for people. We've also found that, honestly, around 24 is when it starts to feel a little crowded. The 32-player number is probably one of the things we'll allow players to do on unofficial bouts. If we figure out a game layout where the 32-player count works perfectly, then we'll do that officially. In the future, we might do four teams of eight or something like that.

How many maps will there be in the early access launch of Beastlink?

For early access, we're launching with three maps. They're big maps, about a mile square, and they're all completely destructible. Our pie-in-the-sky goal is six to eight maps for 1.0. It's possible we'll do less because our focus is entirely on quality.

The Tech Behind the "SuperDestruction" and Server Networking

I know you've injected your own tech, SuperDestruction, into Unreal Engine. Can you explain how it works?

Thomas Williamson: From a very high level, I've always felt that we've got all this consumer hardware and these highly capable GPUs, but we're only using them for frame generation, a little bit of compute, and graphical effects like particles and sprites. There's a lot of stuff being left on the table that's not just basic rasterization.

Instead of having GPUs used only for AI and rendering stuff on screen, we were like: hey, let's harness this power that everybody has for something interesting. That's where the SuperDestruction concept originally came from. We can absolutely drive so much stuff through the GPU if we completely commit to the concept of all of the physics collision and destruction happening GPU-side, alongside all the environmental layout updates. By doing that, we're able to unlock an incredible amount of simultaneous simulation.

As you mentioned, of course, that's very hard to do in multiplayer games because then the destruction has to be synchronized across everyone, too. How did you solve this issue?

Thomas Williamson: Obviously, every game is built a little bit on smoke and mirrors. What we do is try to make sure that the destruction happening on both the client and the server match each other perfectly. There's a lot of work behind the scenes to ensure that when destruction occurs, it happens immediately for each client, but remains exactly synchronized on the server. If they disagree, we made sure there are fast, easy ways for the client state to be fixed.

That implies doing things like creating highly compressed, efficient codecs for communicating that a building is destroyed in a specific way to another client, allowing the client to verify that state before sending data back and forth. In these levels, we've got something like 15 to 20 million separate little pieces that can be destroyed, so we need to make sure that anytime there's a disagreement, the traffic sent back and forth is incredibly small. That was probably the thorniest part of the problem to solve.

What we ended up doing to make that work better is tailoring our game design around it. One thing you'll notice is if you destroy a big chunk of a building and all those pieces fall on you, a lot of times it's not going to do any damage to the player. That was a design choice we had to make because that localized debris collision would be impossible to network with millions of active pieces. However, we can network the structural macro-destruction properly. Conversely, if an entire building starts to collapse and everything is coming down, then you will take damage, because at that point, it becomes structurally important to the gameplay. It's a combination of optimized network compression and very specific game rules about when accuracy between the client and server actually matters.

AI Tools and Steam Content Disclosure

The Steam page of Beastlink includes an AI disclaimer. Can you go into specifics?

Thomas Williamson: Let me read what this description says; I want to make sure I have this right.

Our number one priority when we were making this was not to make an "AI game." Not a single texture that the player ever sees, not a single 3D asset, and not a single animation was made by generative AI. We only put that caveat on the store page because we buy asset packs from the marketplace, as every developer does. Because we buy third-party content, we technically cannot guarantee that a component in those packs wasn't created with AI assistance. I just wanted to cover our ass, but at this point, AI is everywhere, and we are trying our best to ensure everything is handcrafted. Our game doesn't have a lot of historical precursors, so AI wouldn't even do a good job of predicting what we want anyway.

At the same time, I feel really strongly about making sure that the creative art of game development is supported. I really do think we might need to figure out different wording for that store description because a lot of people read it and assume, "Oh, this game's got a lot of AI in it," when it's really quite the opposite. We're just trying to be completely above board. As an example, early on we used generative AI to make placeholder music tracks, but those have since been entirely replaced by real compositions. I don't want to lie and say we didn't try the tools out just to have temporary assets while working on the game.

The most important thing for us is to be completely honest, but people are so skeptical these days that if you are honest at all, they immediately wonder, "Oh, what else are they hiding if they're admitting to this?" It sucks, but that's the way the world works right now.

But personally, and I guess as a studio, what do you think about AI tools like Claude?

Thomas Williamson: It's still early days, and things move fast. Everybody was telling me at GDC this year how Claude was being integrated into their studio pipelines and how great it was. But that had completely shifted over just a month or two. In January, everyone was saying, "Oh no, these tools are so good now, everyone has to use them." I feel like in another few months, there's going to be some other massive difference.

For us, we just keep our eye on it and figure out what's best long-term for the creative art of game development. I have a feeling that long-term, AI will accelerate creative work instead of supplanting it; I'm hoping it accelerates it in a positive way. All these studios closing down or removing junior developers to replace them with AI tools—that's not a great long-term growth decision in my opinion. We're a bit slower on the uptake with AI integration than other studios, and I'm totally okay with that.

Unreal Engine 5, Blueprints, and the Nintendo Switch 2

You have certainly seen Unreal Fest, where Epic unveiled an MCP integration with Unreal Engine What was your take on that?

Thomas Williamson: An official MCP plugin was inevitable. A lot of independent developers were already making custom MCPs for Unreal, so Epic had to make an official alternative. I feel like their hands were tied there; I don't necessarily blame them for making that decision at all. I am curious because a lot of people think Epic is focusing on it more for their scripting tools to push developers towards using AI with scripting layout pipelines, but I'm not entirely sure that's true. We have to wait and see.

Another big announcement at Unreal Fest was Epic's stated goal to slowly phase out Blueprints. How do you feel about that, as a fairly small studio?

Thomas Williamson: We use Blueprints here, but a lot of smaller indie studios are going to be damaged if Blueprints get heavily deprioritized. They've invested a lot of time into understanding Blueprints and building technical prowess around them. I don't think Unreal Engine 5 is going away anytime soon, though—Unreal 5 games are going to be shipping for the next five years easily, so we all have time to sort of figure out what we want to do next. If studios want to pivot away from Unreal for that reason, that's fine. If they want to learn Verse over the course of the next several years, they can do that, but change is inevitable. As a studio, we don't use Blueprints extensively because we don't think it has high execution velocity. I don't entirely disagree with trying to push developers toward a more efficient scripting language, provided the replacement is suitable for entry-level developers trying to get their feet wet.

Epic is trying to sell Verse as something that's, you know, more easily approachable by developers. Do you think that's true?

Thomas Williamson: They are going to try to sell it, though. The jury's out, and we'll find out. I'm along for the ride. Verse as a language is a little cumbersome, but I haven't spent a ton of time in it because this is brand new for me, too.

Can you share what Unreal Engine 5 build you are on right now with Beastlink?

Thomas Williamson: We're currently on Unreal Engine 5.7 right now. Once 5.8 gets a minor stability upgrade, we'll probably upgrade to it. We try to maintain the most recent Unreal version because they typically have the most optimizations.

I think 5.6 and 5.7 introduced many optimizations to Unreal Engine, right?

Thomas Williamson: Yes, and once we got FSR integrated into our Xbox build internally, it stabilized significantly. We're not fully committing to it yet, but we have internal test versions of the game running at 60 FPS on the Xbox Series X. I don't think that would have been possible on Unreal 5.5, but on 5.7, we are able to hit those frame rates.

Do you think Beastlink could run on the Nintendo Switch 2?

Thomas Williamson: If the game is successful, I'll certainly try. It depends on the success level, really. I love working on smaller hardware—it's no secret if you look at our porting history. I do think it's possible. It won't look exactly the same as the PC edition, but I think we can get it 75% to 80% of the way there so people can have that portable experience. But I don't want to commit to it until we know we have the available development bandwidth.

A lot of people believe that Nintendo was kind of helped by the fact that, you know, there is the Xbox Series S, so developers already have to optimize for that platform, and it's not that much more powerful than the Switch 2.

Thomas Williamson: Yes and no. The only problem is there are still some structural constraints on that hardware not related to sheer rasterization and compute power. The Switch 2's biggest hurdle is that it is really difficult for Nanite to run on. Nanite is fundamental to a lot of modern Unreal Engine 5 games. Until Epic or Nintendo solves the Nanite issue on mobile chipsets, it's going to be highly problematic for developers to port to it.

I anticipate you'll see far fewer third-party Unreal Engine ports on the Nintendo Switch 2 unless that optimization issue gets solved. It's concerning because Epic didn't announce a lot of Nanite pipeline improvements that would make that feasible with version 5.8, and they've stated that 5.8 is likely their final major iteration of Unreal Engine 5. I'm concerned it might really limit the potential of the Switch 2 to receive third-party Unreal-based ports.

Campaign Mode, Release Window, and Pricing

What's your current early access launch window for Beastlink? And also, can you share something about the price?

Thomas Williamson: It's coming out this year, I can say that for sure. And it's not going to be a $60 game, I can tell you that. It's an indie title, and we're going to price it as an indie title.

Did you ever consider adding a campaign mode to the game if there is enough demand?

We have a complete story and lore built up. For the campaign, we are going to include a single-player mode that allows you to play through levels that unlock different items. However, they are essentially glorified scenarios that utilize the multiplayer framework. In terms of having a fully-fledged narrative campaign with dedicated cinematics and all that production value, that's something best left for Beastlink 2 one day.

Thank you for your time.

Follow Wccftech on Google to get more of our news coverage in your feeds.