One developer. A custom engine. A virus that turns humanity’s own technology against it.
We spoke to Korányi Levente, the independent creator behind Fleshgrinder, about his love of Diablo II, the technical challenge of enormous enemy hordes and his plans for this dystopian action RPG.
The questions below are those sent by Coloured Static Press. Levente’s answers are reproduced with light spelling, punctuation and formatting edits for readability. Gameplay and development claims are the developer’s, not independent hands-on verification.
Let’s start at the beginning. Fleshgrinder brings together the loot-driven gameplay of classic action RPGs and a particularly grim vision of humanity’s future. Where did the original idea come from, and at what point did you realise it could become a full game?
Okay so since I was 4, back in ’94 I got completely hooked on computer games, started playing Doom and Wolf3D with the help of my stepfather who taught me how to start them up. By the time I was 10–11, I could put together PCs and got familiar with Fallout 2 and Diablo 2, which completely blew me away. I was looking for the same experience to relive but no other game came close to them, especially Diablo 2 which I sometimes still play, even on my phone.
Given my history with tech and computers, I try to keep up with the latest and greatest, and once Neuralink came to my attention, ideas began forming in my head about the future. Humanity is now capable of controlling computer systems with a Bluetooth chip that gets implanted as a BCI? I mean WTF, how sci-fi is that? Many people still won’t believe me when I tell them about it.
So I started wondering: what if this technology advances so much that information transmission stops being one-way, so thoughts could be implanted, giving way to practically nasty stuff as well, like a digital virus that could infect minds? Hence, the basis of the story of Fleshgrinder was born. That was two and a half years ago.
Since then, I’ve tried Godot, GameMaker Studio, but none of those GUIs and engines was appealing to me and given that I’m not much of a programmer but AI coding agents are now advanced enough to be able to put together sophisticated software, I thought I’d create my own engine from scratch.
Started as a weekend project with a tower-defence game in mind initially, using Pygame and nothing but CPU rendering from a top-down viewpoint. As time passed, I got more and more hooked on creating ever more sophisticated systems for the game and started transforming it into GPU-driven rendering using OpenGL and moved other CPU-heavy parts of the game from Python to C++, creating this hybrid monster it is today.
Converted the viewpoint from top-down to isometric and since I still think that Diablo 2 is the greatest game ever made, I took inspiration from that as well as some movies like 28 Days Later since I’m also a big fan of zombie movies and games.
The story explores a fascinating idea: technology designed to improve humanity ultimately becomes the thing threatening its existence. With SynTech, the Convergence initiative and the Fleshgrinder virus, there’s clearly more going on than a traditional zombie apocalypse. Are you deliberately exploring our increasing dependence on technology, or did those themes emerge naturally as the story developed?
It was formed around the time when I started thinking about Neuralink. The story evolved over time but the base idea is pretty much two and a half years old. Second was to invent and figure out the characters and their roles in a collapsed society, then came the factions.
I’m still not satisfied how much background lore there is and I will definitely introduce more factions and companies to better represent the world I have in mind for the game.
SynTech represents one of the megacorps that ruled the world, a bit like how today’s big tech does, just in a more advanced way, since our increasing reliance on tech and the rich getting richer while the poor get poorer points to such a possible dystopian cyberpunk-ish world already.
One of the things that stood out to us was the distinction between the Purists, the Disconnected and the Immunes. These are three very different responses to the same catastrophe. How much will those differences influence the player’s relationships, decisions and experience of the world?
They are unfinished as of now. What I have in mind for them is that every human still alive will be part of one of the three backgrounds. I have some fanatic religious cults in mind as Purists; for the Disconnected, they are usually crippled and weakened people, downtrodden of a ruined world, saved by Gutters who did everything they could to remove SynTech augs before they could take over; and the Immunes, who got lucky enough and resisted or avoided the infection through manufacturing irregularities, or illegal modifications or malfunctions.
One of the NPCs, Ralph, for example, got himself electrocuted working on one of SynTech’s facilities’ electric fences that fried his implant right before the virus hit. He was lucky enough that the wireless link broke.
I have these backgrounds in mind for the characters which the players can choose when creating them. Then it will modify certain NPCs’ reactions and unlock certain dialogue options, quest lines, or be completely dismissed and maybe even hostile by some. Augmentation slot numbers are also varied between classes, and choosing a Purist, for example, would lock all, while giving bonuses to other stats.
Fleshgrinder’s three operative classes, the Butcher, Hacksmith and Mindripper, have some wonderfully distinctive personalities and abilities. How did you settle on these particular characters, and have players or testers discovered any unexpected combinations of abilities that surprised even you?
When I was thinking and imagining the game world it became apparent and natural that there will be specialised roles for the remaining survivors. A grunt/brute who serves as a tank, a healer/support who can also be a ‘mage’ type of psychic and a hacker/tech specialist with gadgets and knowledge about electronics and mechanics came pretty soon naturally.
There isn’t too much feedback yet. They are all very distinct and the balance will still need more fine-tuning especially towards endgame, so it’s a little bit too early to tell yet.
The idea was to have a squad of specialists instead of a one-man army, but hiring more operatives yields reduced XP to them, so I’m trying to leave the opportunity for completing the game with just one character as well.
The Genesis Tier system is particularly intriguing because the virus can become progressively more dangerous. How important is that escalating threat to the gameplay experience? Do you want players to feel genuinely pressured into making difficult choices rather than simply clearing every area and collecting everything?
Minor story spoilers follow.
There are events in the game which are unavoidable through the story which will advance the virus and the Genesis Tiers.
I came up with the idea to have the monsters ever increasing and engaging them still dangerous on later missions when the player is strong enough to blast through them, so not only more advanced entities are to be faced, but the original, weaker ones like the Sprinters and Zombies are also getting a little bit more aggressive.
This system also came naturally as the original idea was that the virus collects materials to advance itself and since it has a wireless link as a hive mind to all the infected, it keeps evolving and thus, like a software upgrade, can strengthen them.
There are areas which are extremely dangerous if entered too early as some facilities have advanced Fleshgrinder operations in place which start when the player arrives. Those areas usually contain infrastructure the virus uses for the collection and if the player isn’t fast enough to stop it, the advancing of the Genesis Tiers will most certainly kill them.
The demo already has this mechanic and I’ve made it so players can let the machines work, collecting materials to try to survive against the virus as much as possible. The mission Unhappy Landing, which is the first big one, is resettable, keeping character and gear, allowing for endless farming.
There’s an impressive amount of interaction within the game world, from barricading doors to hacking terminals and dealing with enemies that can pursue players between areas. Was creating that sense of an interconnected, persistent world always part of the plan? And have any particularly memorable or unexpected moments emerged during testing?
A persistent world was always the plan and building multiple missions from the start already had this implemented. A truly open world would be the absolute best, but that’s not going to happen. There are pretty huge maps/missions already and a couple of seconds of loading between them is acceptable to me. The development is way too advanced to start tearing the engine apart and try to make it into an open-world game.
The game has many events that are tracked via world flags. These are knobs that something was already done or not so that systems which rely on them can fire at given moments, i.e. ‘If terminal X has been interacted with then NPC X unlocks this dialogue option’, allowing independent events to be tracked across the entire game world. This also allows for interesting story-altering choices and side quests unlocking, for example.
It was also pretty challenging what to do with pursuing hordes, since they can follow the player throughout missions. I had to think of a way that is practical and manageable. For example there is a 10-second delay between when the player can travel between missions and there are a couple of seconds when they can get ahead of a pursuing horde. They are also aware which mission the player arrived from and will start spawning from that particular mission’s exit, still chasing. Hordes can also merge, as a global counter keeps track of its members dynamically.
There’s also a delicious irony in a game where cybernetic augmentation helped destroy civilisation, yet players can still enhance their own characters with implants. How do you balance the temptation of becoming more powerful against the dangers and consequences associated with those upgrades?
Gutters are crafty enough that they could separate the network stack of these implants so the ones they are required to implant are safe.
Most implants also have downsides, while giving great bonuses. For example, the Convergence M-graft is an Endo-aug, which does not require Gutter service to be implanted. It significantly strengthens a character as an external reinforced spine. While being robust, it decreases the agility of the wearer. A Neuralense needs implanting with the use of a Gutter and NexuGel. It gives sort of a tunnel vision, giving enhanced view range and accuracy but decreasing field of view and vitality.
You’ve been developing Fleshgrinder independently, and the scope is ambitious. What’s been the single biggest challenge you’ve faced so far? Equally, is there a particular feature, technical achievement or small detail you’re especially proud of that players might otherwise overlook?
Making such a sophisticated game in two years which can handle 1,500 monsters at once on a map, and having them properly pathfind inside houses from across thousands of pixels and mazes was extremely challenging and took me multiple tries and weeks to get it more or less right. I’m still not a hundred per cent satisfied with it since every now and then an enemy might still get stuck in a doorway but the performance is pretty good and pretty much every time I touch it it just becomes worse.
Combine this with penetration-based mechanics and multiple miniguns, rocket launchers mowing down hundreds in a matter of seconds without blowing up PCs and without a licensed engine that does this out of the box is something I’m proud of.
Since everything is in-house developed, I have full control over it all and can improve on anything at any time, which allows me to keep developing the game for as long as I’d like to. As you said, the scope and features are very numerous so there are many more I could list.
With the playable demo available and Fleshgrinder targeting an early 2027 release, this is an important period for the project. What has player feedback taught you so far? Has anything prompted you to reconsider an original design decision, and what are your biggest priorities between now and release?
Since Act 1 is more or less done, 14 missions are playable but there are many more to come, so the campaign has to be finished and I’ve yet to explore all story options to wire in-game. There is still plenty to do besides that as well: NPC factions are still extremely underutilised, dialogues need more options and better wording at certain places, terminals could do with more lore and interesting files, some monsters still need polishing and the performance is something I’m always looking to enhance as much as possible so weaker hardware can also run the game.
Player feedback was very important so far since when I get to see others play who don’t know the game it always gives me a ton of insight into what to modify and polish so it’s easier to understand.
The tutorial, for example, was way too long and I’ve cut it in at least half and decided to implement a hint system in the campaign instead which pops up on certain events to explain a feature instead of squeezing everything in the tutorial.
Finally, we’d love to finish on something a little more personal. After all the time, effort and creativity you’ve invested in Fleshgrinder, what do you most hope players take away from the experience? Is there something you particularly want them to remember long after they’ve stopped playing?
I’m aiming for a unique and terrifying experience which is relying on the overall atmosphere and the deep storyline. ARPG with realistic guns and gory body horror which utilises a story that is not so far-fetched from reality is uncommon and I hope that I can bring the same great experiences I could live through playing with my favourite games.
The systems allow for infinite farming and strengthening of characters, finding awesome loot where premade uniques all behave and give something on their own which no other item does. That dopamine hit was one of the most important hooks in Diablo 2 and as such, I’m hoping to recreate it for the players.
One more thing: how the bullets work
Levente also sent us an additional technical note on the game’s penetration mechanics:
Korányi Levente: The penetration mechanics are also more or less real-life based and ammunition like the 7.62mm will easily penetrate wooden doors and possibly kill beyond them, allowing for threats to be taken down without them being able to get to the player. Every material has a value for how much penetration force is needed to be able to shoot through it and every ammo type has a value for how much it can penetrate.
Weapons can also roll with penetration bonuses. Line up five zombies for a sniper shot to take them all down with just one bullet. Blast between them with a shotgun and some of them might just get knocked down with an arm dismembered, which will then still get up and leave a blood trail, still coming at you while you reload.
Editor’s note: Fleshgrinder remains in development. The early 2027 release window referenced in our question is a target, not a confirmed launch date. Some described features are unfinished or planned.
Our thanks to Korányi Levente for taking the time to answer our questions. Discover more independent games and developer stories at Indie Discovery.