For the first anniversary of Baldur's Gate 3, we talk to Larian about how it reinvented itself, why it scrapped its plans for Baldur's Gate 4, and what the future holds

It's early July, almost a year since Larian launched Baldur's Gate 3—a year in which the studio has collected a vast number of awards and accolades for its monumental RPG—and the team has swapped the Forgotten Realms for the swelteringly hot Spanish city of Barcelona. Larian is here to discuss its next game behind closed doors. A couple of days ago, on July 8, it held a summit to talk about kicking off a new RPG. "Lift off," as CEO Swen Vincke puts it. 

There's the crackle of electricity in the air, not because a dragonborn sorcerer has just cast chain lightning, but because the team is about to set off on a brand new adventure. Larian's developers are now spread all over, including Dublin, Ghent, Warsaw, and here in Barcelona, where so many of them have gathered. Much of the excitement, then, is also down to a lot of them being in one place, rather than distributed across the world.

It wasn't all that long ago that Larian was comparatively tiny. "These are guesstimates," says Vincke, "but I think that Divinity: Original Sin 1 must have been around 40 people, and then Divinity: Original Sin 2 was around 120-ish. And now we're approaching 500." The new Warsaw studio has added a chunk, and Vincke says that it's "beefed up" QA, adding more internal playtesters to prepare for what's coming next.

So what is coming next? Well, that's a secret that Larian's not quite ready to share. Chatting to members of the studio following the summit, however, reveals a lot, from hints about the scope of its next two projects to why exactly the team stopped working on what was shaping up to become Baldur's Gate 4.

Little big developer 

Years ago, I visited Larian's Ghent studio for the conclusion of D:OS2's successful Kickstarter, and while it had grown considerably after D:OS1, it still had this small studio energy. Writing director Adam Smith, who joined Larian for Baldur's Gate 3 after years in the games journalism trenches, reckons that energy continues to be maintained despite the huge growth. 

"I was a big fan of Larian, and my biggest fear joining Larian was growth might make it not the studio that I saw from when I'd spoken to Swen and other people in the past. It hasn't changed. I mean, it's crazy. The energy when people talk about making things feels like a small company. Actually, somebody said something today, one of our producers, that game development is a bunch of people sitting around a screen getting excited. And that's still how it is."

"I call it chaos at scale," Vincke adds. "We're a studio that gives a lot of freedom to developers. We encourage it, actually, bottom up, because it's unmanageable otherwise. Different people have to do their own thing. And then obviously, we have to manage the chaos that comes out of it, so that's a large part of the job. But the fact that we start like that, and we have been on purpose keeping studios not too large—so they're basically people that all know each other still, which is a very important part—is the thing that makes it work. It's as simple as that. So it is just empowering developers themselves."

An especially saucy scene from BG3 is a good example of this. Smith was walking through the mostly empty office when he spotted something the cinematic animator and cinematic artist had been working on. It was the final part of the now infamous Halsin bear sex scene, when we see a squirrel react to it by dropping its nut. "That was them independently just saying 'What's a cool way we can end this scene?' And they experimented with it, they tried it, we happened to see it, and we were like 'Holy shit'. [We] didn't know at the time that was going to become famous."

This structure of flexibility and experimentation conveniently matches Larian's design philosophy, which Vincke describes as "player first, always". And this has been evident across all of its games, where player agency is paramount. "I started my summit here by literally telling [Larian], treat the player like you would want to be treated. That's literally how you should make a game. I keep on repeating, if you see something in the game that you would not want to see in another game, don't do it—just keep on working on it, or don't do it at all." Larian's systems all stem from that rule, which is why its RPGs don't feel prescriptive. "If it's on rails, you can go to a theme park; this is not a theme park."

Giving players access to powerful spells like teleportation or talking to corpses early on really encapsulates this philosophy. In BG3, you have so many ways to get out of a scrape, uncover hidden areas and get up to all sorts of mischief right away, rather than having to wait 50 hours before you can start truly doing what you want. But this also means Larian's games have to be designed in such a way that the developers must get into the players' heads and figure out what they might want to do. 

"We have to foresee a whole bunch of things," says Vincke, "and then we have to make sure there's a fallback for all the things players will come up with." And then there needs to be all these rewards and motivators to keep players exploring. Fancy items, compelling storylines, big surprises. "And that's essentially how we approach games. "So everything we do, every single thing within the studio, is built to allow us to do more of that."

A publisher's nightmare 

One of the reasons Larian can do this is because it isn't beholden to anyone else. It's an independent studio that doesn't need to rely on publishers, and while it does have investment from the likes of Tencent, they don't have a seat at the table.   

"If you see it, it'll scare you, as a shareholder, as a publisher," says Vincke. "The very first thing you will try to do is get it under control—you will try to scope it, you will try to box it, you will try to control it, which is exactly the opposite of what we want to do. These days, especially when you have so much entertainment, people are only going to put their time into something that keeps on engaging them. In the type of game that we make, that means that we need to have a lot of content, and you put a lot of content in there that never nobody's ever going to see. Which is really not obvious, right? Because it's like, why are we spending a million dollars on a dragon, which nobody's ever going to see, except like five people who made that obscure choice? Because if they see it, they have to be happy too. So that's why I put the bloody dragon there, because that's the logical consequence of the things that they've done."

While Larian's approach would likely terrify a lot of publishers and investors, there's no doubt that many of them would still like a piece of this pie. The industry is obsessed with the studio, and all eyes are on what it's planning next. In the year since launch, it has undeniably become one of the world's most influential developers. But Vincke attributes its series of big successes, from D:OS onwards, to its biggest failure: Ego Draconis.

Divinity 2: Ego Draconis was the third game in the Divinity series, following on from Divine Divinity and Beyond Divinity. Vincke recalls that it had a lot of things that could have made it a "breakthrough game", but unfortunately it didn't come together.

"It was ambitious, had a lot of stuff going on, but none of it was really well done," he said. "So everything was really the product of being too ambitious with too small a team and not having enough resources, and then the publisher insisted on releasing it when it wasn't ready. And so I fought very long and hard about that, because I thought about quitting at that point. You can't just go from failure to failure to failure. You've got to do something right. This is famously when I dropped publishers."

Vincke still believes the game had a lot going for it, like the combat and the narrative, even if the latter "didn't deliver". But Larian had attempted to do too much. "We couldn't do all these things because we were too small. So we needed to reinvent ourselves. I said, 'If we do one thing right, it has to be the combat, and you have to be able to play it with your friends.' And that was D:OS1. The story was really tacked on. D:OS1 was a success, so I looked at that and said 'Now we need to nail the story.' D:OS1's story was done in like three days or something. Now we spend years discussing story and trying to develop characters."

As well as this reinvention came things like the complete dominance of digital distribution, "and there weren't 320,000 middlemen that were just taking cuts on everything", as well as the rise of Kickstarter and early access. This helped Larian fund D:OS1 and 2, and then actually make money out of them. Now, though, it's in a position where it doesn't need Kickstarter. Unsurprisingly, BG3 has filled its coffers. 

"We launched, and we were nervous as hell, " Smith remembers. "Early access helped with that, because you knew that people enjoyed what you had, but it was still really nerve wracking. And I remember at some point in the night, we were having a celebration. We were looking at the Steam numbers, seeing the numbers going up and up and up. And the first time I spoke to Swen after launch, I said something like 'Did you expect this?' And he said, 'You know what this means? It means we get to do it again.' That was really it."

The summer of RPGs 

BG3 launched during a busy summer for RPGs. Diablo 4 had just arrived, and both Starfield and Cyberpunk 2077: Phantom Liberty were imminent. Larian shifted the release date to avoid a clash. "Big comes before small," says Vincke. "The biggest release gets the space, and then the others have to move for it. This has always been a universal truth. So we recognised Starfield as a bigger game than us. So that's why we had to move. Players were not going to be playing multiple RPGs of this size at the same time. That's impossible. So you were going to force players to make a choice, so we said 'OK, we'll take that choice away and move elsewhere.'"

Instead of Starfield sucking all the oxygen out of the room when it launched, however, BG3 kept selling, and continues to. Hardly a day goes by without sites like ours covering it, and you all keep reading about it. You're ravenous, even a year after launch. For a game that isn't live service, which isn't going to be getting any DLC or expansions, it's almost unheard of. Larian ended up going so much further than just managing to compete with Bethesda, Blizzard and CDPR.

Since launch, Larian has released a large number of patches and hotfixes, squashing bugs as well as adding new features, an epilogue, more evil options—countless additions, tweaks and treats. But now Baldur's Gate 3 is done, for the most part. 

"A complete experience," says Vincke. "There's lots of stuff we wanted to do that we ended up not doing. People always ask what are the features that you didn't want to cut. But in fairness, I think you got what you paid for when you start playing the game. You got a good experience. You had a lot of choices, you were rewarded for those choices, the game respected them in many, many ways. You got a good character development experience. You had fun in combat—lots of tactical combat that you can approach in many, many different ways. These were the things we wanted to create. We're moving on, so in all our hearts the game is done."

"People will datamine things that are in the game," adds Smith, "but there's a reason we didn't put them live, because we felt they detracted. Putting more in and continuously tweaking can make something worse."

Larian has been discussing BG3 for years, and working on it for a lot longer, so while Vincke and Smith still talk about it with enthusiasm and eagerly share anecdotes about its development, it's tinged with, if not exhaustion, then at least a readiness to be done with the game. BG3 was initially planned to be a one-shot. No expansions. No sequels. At least not from Larian. But there was a period when more Baldur's Gate was on the cards.

Baldur's gaol 

"So if something is super successful, obviously everyone's [asking] you, 'When are you going to make the next one?'", says Vincke. "And when you, as a developer, come out of this little miserable cave that you've been sitting in for many, many hours as you're finishing your thing, you're vulnerable. You're emotionally drained, because you don't know what the reaction of the people is going to be; there's so much hope and anticipation in there, there's a lot of responsibility. So you tend to be prone to do the obvious thing, which was really just make an add-on, or a standalone add-on, or start working on a sequel, because it's the easiest route to take. And we actually started doing that, because we've got the engine, got everything in here, all we need to do is add extra levels or just make some new stories—how hard can it be?

"It's something that you all would have liked, I think. I'm sure, actually. And we actually went pretty fast, because the production machine was still warm. You could already play stuff. But you played it and you looked and it, and, like, you know, this is OK. I mean, we'll probably have to redo it 10 times. And do we really want to do this for the next three years? Because we thought we'd make it in one year, and then you could see it's not going to happen, it's going to take two, three years. And having just done six years of D&D, which is not our own thing, are we really going to spend all this time on this and abandon our own plans? Yeah, maybe not for an add-on, but maybe BG4? That makes a lot of sense. So why don't we do it for BG4? Oh, yeah, that sounds like a really good idea. Let's make BG4. All the stuff that we did for this thing, we can just move it in there, people are gonna love it. Go for it. And then the same thought process again."

At this point, Vincke dramatically sighs, reliving the moment he decided he didn't want to be trapped in a D&D-shaped prison. "Another X amount of years, same thing, exactly the same mechanics, same problems that we actually already solved, just evolving them. And then you start looking at the other things that you had planned and pushing them back, and then you've come back to your senses, because by then the vulnerability is gone."

Staring ahead at this less than ideal future, Vincke stopped thinking about the easiest route. "We should be looking at how we can do stuff that we get excited about." After a long overdue break during Christmas, he had a reality check, and then talked to the team about the prospect of moving on.

"It very rapidly turned, and I don't think, as developers, we ever felt better since we took that decision," he says. "Honestly, you really cannot explain or express it, how liberated we are. So morale is super high, just because we're doing new stuff again. We're doing our own thing again, we're not rehashing, we're not trying to convert rules from 50 years ago into something new."

Vincke and Smith emphasise that it was not so much an issue with D&D—"It will always be one of the coolest things I ever did," says Smith—but that Larian was tired and needed something new to energise it again.

"It needs to get the spark," says Vincke. "What we did here at the summit was pitch to our own teams. 'So this is the stuff that we're doing, are you excited?' It's the question everyone was asking: are you excited? You have to have that.

Doubling down

While one is further along, the plan is to develop two new games concurrently, but that's already proving to be a challenge. "We had some people who were dedicated to the other one, and they just got swallowed by the ambitions of the one that we want to release first," says Vincke. "So we will try to do it concurrently, but we have to figure out how to leave people alone." 

With two games on the slate, one might imagine that Larian's going to rein in its ambitions, at least when it comes to each individual game. But that's not a very Larian thing to do. "We always say we do," says Vincke, "but then we start thinking, and the machine is meant to make large games. We were made for making large, ambitious RPGs and to try new stuff. One of the core problems—it's not really a problem—but one of the core problems to do what you're suggesting is that we actually know what we want from our gameplay systems, how to evolve them, how to do new things … and they're all big."

One of the tricky parts of following up a very big game with more big games is trying to find new things to fill them with. "One of the biggest problems we have now is that whenever we're talking about things, we say we did that in BG3," says Smith. "And it turns out, we did a lot of things in BG3 when we think back to it. But it was the same during development. Have they already seen this pattern? Have they already used these verbs in this order? Have they already had this emotional arc? So you're constantly trying to make sure that they're getting a new experience, and you're not just repeating yourself, and you're not just giving them content for the sake of content."

It's not scale for the sake of scale, either. A short RPG can be a great RPG, like Disco Elysium, but as Smith points out it's a very different kind of RPG from the ones that Larian makes. "[Disco Elysium] is a case. It's still a journey, but it's much more contained. And it's a very different psychological journey. In a party-based game, especially in a multiplayer party-based game, you need people to be able to go off and do their own thing for a while, discover their own stories. Discovery is really important, and for discovery you need space, like you need the moment where they explore and go 'Holy shit, I just found something completely different over here," and for that you need the length and the space."

Even with the massive size of BG3, Larian still struggled to find space for all the stories it wanted to tell. It was actually envisioned as larger, initially. Wyll had his own area, the Red War College. Instead of meeting Gortash in Baldur's Gate, players would meet him in Candlekeep, where the first Baldur's Gate began. There was a Shadow Druids hangout, a gnome village, and a map that could be explored in a non-linear fashion just like the original games. These were cut for a variety of reasons, but one of the key issues was that they were smaller areas, leading to fewer ways for players to be able to express themselves and do their own thing. Larian opted for fewer but larger areas where more things could happen. But that also necessitated finding new homes for some characters and story beats.

On the subject of new homes, is Larian interested in leaving its well-trodden fantasy realms behind? Both Rivellon and the Forgotten Realms share a great deal, and Vincke and Co have been dabbling in this sandbox for decades. "Yes, of course," Vincke says. "There's so many things that I would like to do. Just so [few] things that you can do. So there's really a lot I would like to do…" And not for the first time during this interview Vincke catches himself about to give too much away, and clamps up. He won't be falling into any of my brilliant traps, it seems.

IP bingo

While Larian is back doing its own thing, it's hard not to hope that it will eventually take on another existing property, maybe something that's been languishing for years. One series Vincke often discusses is Ultima, and in many ways the Divinity games are love letters to Richard Garriott's classic RPGs. Indeed, when Vincke was considering properties for Larian to work on, before it got its hands on D&D, Ultima was on the list, along with Fallout. Broaching the subject now, however, he seems to be mostly over the idea. 

"That's Richard Garriott's thing," he says. 

"I love Ultima," adds Smith, "but I don't think it needs to come back. The thing about Ultima for me was, 4, 5 and 6 I absolutely love, but 7, it might be my favourite game. But I don't know. Like, would you remake Ultima 7? I don't know what you'd want to do. With Baldur's Gate, there was a world I was interested in. There were characters and there was a story that felt like they could continue, and themes that were very strong. Ultima, each game … they're very independently themed."

The most interesting component, the thing worth exploring, Smith says, is the virtue system. He'd be interested in going back to the earlier Ultimas, "where it's like, how do you make an RPG that treats combat as a violent thing that is not good, which is one of Garriott's original ideas. You make RPGs or games where you hit things on the head and looted them. And he said, 'Well, what if, actually, that is a bad thing to do. What if you have to try and find peaceful solutions?' That's really interesting, but that happens across the board in RPGs now, so it's not as interesting a question as it was. So yeah, it's not a rich world to me."

Despite this, Vincke isn't closing the door entirely—to Ultima or another property. "Never say never," he says. "There's another developer who I'd love to work with one day in my life and we'd get along really well. I can't tell you who, but he's a big Ultima fan, so you never know."

What's incredibly clear, though, is that Vincke and Smith are much more excited to be free to create things that are uniquely theirs, playing in sandboxes of their own creation without the constraints of a ruleset that somebody else conjured up. And while, yes, as a D&D fan it is a shame to see Larian leave Baldur's Gate behind, it's actually thrilling to know that the best RPG developers around are now unshackled, free to take what they achieved in BG3 and push it forward or in other directions without limitations—other than self-imposed ones. For the last 10 years, Larian's trajectory has been 'bigger and better', so its next adventure could be its most impressive. For now, though, only Larian knows what it's going to be.