Brotato should not have been a Godot success story. A free open-source engine with no royalties team, a solo developer on his fourth small game, and a potato that holds six guns at once sounds like a game jam joke — yet French developer Thomas Gervraud turned it into one of the most popular roguelites on Steam in 2022 and 2023, with overwhelmingly positive reviews and over one million copies sold during early access alone.
The short answer for developers is this: Brotato was built in Godot, listed as its engine, by one-man studio Blobfish, and shipped first as Steam Early Access for Windows on September 27, 2022 before its full release on June 23, 2023. Godot's boring 2D competence — tiny builds, fast GDScript iteration, and painless PC-to-mobile export — fit a 30-minute auto-shooter far better than Unity or Unreal would have. That fit is the lesson.
This article explains what Godot actually is, why it fit Brotato's wave-and-shop loop, and what I think most coverage misses about boring engines winning.
Who this is for and what is covered
You are a solo or small-team developer choosing between Godot and Unity for a 2D systems game, or a technically curious player wondering why Brotato feels so light and runs almost anywhere.
Scope as verified in October 2026: Brotato is a 2023 top-down shoot 'em up roguelike developed by French independent developer Thomas Gervraud under Blobfish, described as a one-man indie studio based in Lyon during development and the game's fourth title. It entered Early Access on Steam for Windows on September 27, 2022, with full release on June 23, 2023. Android and iOS ports by Erabit Studios followed on March 28, 2023, Switch on August 3, 2023, and PS4, PS5, Xbox One and Series X/S on January 30, 2024, with the Xbox versions on Game Pass at launch. Its first DLC Abyssal Terrors plus local co-op arrived in October 2024. All dates and platform splits are from the Brotato overview.
Gameplay in one paragraph from Steam: a spaceship from Potato World crashes, sole survivor Brotato can wield 6 weapons at once, you survive hostile waves, collect materials, and buy upgrades between waves. Steam lists auto-firing weapons by default with manual aim option, fast runs under 30 minutes, waves lasting 20 to 90 seconds, dozens of characters and hundreds of items and weapons from flamethrowers and SMGs to sticks and stones, plus local co-op for up to 4 players. Minimum PC spec is strikingly low: 2 Ghz processor, 4 GB RAM, 128 MB graphics with OpenGL 3+, and only 200 MB storage.
What Godot actually is — and how niche it is
Godot is a free, open-source 2D and 3D game engine, released under the MIT License and initially developed in Buenos Aires by Juan Linietsky and Ariel Manzur before its public release in 2014, as summarized in the Godot engine overview. It supports C++, C# and GDScript, uses a hierarchy of nodes organized into reusable scenes connected by signals, and stores all resources as files on disk rather than in a database to ease version control.
How niche? The editor itself is only 28–189 MB depending on OS, versus gigabyte-scale Unity or Unreal installs with console SDKs and lightmappers. It enables PC, mobile, web, console and XR export without licensing fees. The official site pitches it plainly as your free, open-source game engine for 2D and 3D, cross-platform or even XR — no royalty calculator, no seat audit.
That is the niche-plus-popular contrast: a community-owned engine with no sales team powering a Steam hit that sat next to Unity games on the most-played roguelite charts. Godot's own 2022 retrospective explicitly names Brotato alongside Dome Keeper and The Case of the Golden Idol as hit games published with Godot that year, when Godot 4.0 was still in alpha and beta stabilization.
Search demand proves the niche is real but underserved. Google suggests brotato godot engine as a completion, and Bing returns a correlated Brotato result set for queries like Brotato developed Godot interview Blobfish, while broader Slay the Spire libGDX style queries collapse into dictionary noise. Players and devs ask what engine the potato runs on, but few pages answer it with technical depth. This article fills that gap.
Why GDScript plus Godot fit a one-person arena roguelite
Verified history: Blobfish is a one-man operation, Brotato is its fourth game, and early access moved fast: September 2022 early access, March 2023 mobile, June 2023 1.0, all while Godot 4.0 missed its 2022 stable target and aimed for early 2023. Shipping during that turbulence on a mature stable line was pragmatic, not ideological.
My reading of why the stack was rational:
Iteration speed beats rendering features. Brotato lives or dies on shop balance: dozens of characters with different starting stats, hundreds of weapons and items, materials-for-XP economy, 20 to 90 second waves. That is hundreds of number tweaks, not shader work. GDScript's tight scene-script loop plus file-system scenes means a solo dev can change a weapon, press play, and test a 90-second wave in under a minute. No lightmap bake, no domain reload roulette.
Auto-fire removes what a small engine is bad at. Like Vampire Survivors, Brotato auto-fires by default. You move, position, collect materials, and choose builds between waves. No precise twin-stick prediction, no rollback netcode, no 3D occlusion — exactly what Godot 2D batching does well. The 6-weapon cap looks like power fantasy but functions as readability budget: six arcs plus 100 aliens is already visual noise.
200 MB is a strategy, not an accident. Steam's 200 MB install, OpenGL 3+ and 2 Ghz minimum let Erabit port to free-to-play and premium mobile versions without a rewrite. Contrast Vampire Survivors, which started on Phaser JavaScript and had to rewrite to Unity at version 1.6 for performance and consoles. Brotato never needed a rewrite because it asked so little from the start.
The design-engine fit most coverage misses
My analysis — labeled as opinion: Most roundups say Brotato succeeded despite Godot, or because Godot is free. I think it succeeded because Godot is boring for this genre.
Short waves plus shop pauses act as a garbage-collection shield. GDScript, like JavaScript, is garbage-collected. Vampire Survivors' 30-minute continuous runs exposed Phaser's hitches exactly when 2,000 skeletons filled the screen. Brotato's 20 to 90 second waves followed by a full-stop shop screen give the runtime a natural breathing point to clean up bullets, damage numbers and pickups. The loop hides the engine's weakest moment behind the game's most fun decision. Whether intentional or not, that is engine-aware design.
The potato joke is scope control. One potato, one arena, top-down sprites, no story cutscenes, no voice acting, runs under 30 minutes. Every asset you do not need is a Godot subsystem you do not have to learn. A Unity prototype with free NavMesh, Mecanim and VFX Graph tempts you to add 3D, longer runs and smarter AI — and then you are no longer shipping in nine months from early access to 1.0.
Provocation 1: Would Gervraud have shipped Brotato in Unity without doubling scope? I doubt it. Abundance invites complexity. Godot's modest 3D in 2022 was a feature: it told a 2D designer to stay 2D. What would your prototype look like if you were forbidden from adding rendering features and only allowed to add shop items?
Provocation 2: The fourth-game effect matters more than the engine. Brotato is Blobfish's fourth title. File-system scenes, signals and GDScript are learn-once tools that compound across games. The engine did not make the designer fast; the designer reusing the same boring tools made the engine look fast. Are we measuring engines when we should measure repetitions?
Open questions I cannot resolve from public sources: exact Godot 3.x versus 4.x line used at early-access launch, object-pooling strategy for late-wave aliens, and how much mobile performance work Erabit did versus upstream Blobfish. Those need a postmortem or code access. I leave them open rather than guessing.
Feel the loops in your browser before you commit
If you want to understand Brotato without installing 200 MB, practice its two core skills in free browser games.
To feel wave positioning under pressure, try Iron Snout. Its Classic mode is endless survival against continuous wolf waves where one mistimed move ends the run, with Sudden Death for flawless-play practice and Wolfieball for local 2-player chaos. It teaches the same centered positioning and pattern recognition Brotato demands in 20 to 90 second bursts, just with hooves instead of SMGs.
To feel weapon-progression tinkering, try Stick Merge. You merge identical pistols into submachine guns and assault rifles, balance active-slot power against merge timing, and chase damage scaling — a distilled version of Brotato's between-wave shop where you decide whether to merge now or save materials for a higher tier. Both run instantly in the browser, like Brotato's promise of fast runs with no setup friction.
What to steal (and what not to copy)
Steal this if you are solo:
- Choose the boring stable line during engine turbulence. Brotato shipped September 2022 while Godot 4.0 was still in beta; stability shipped the game.
- Design waves around GC pauses. Short combat plus full-stop shop masks hitches better than any profiler trick.
- Cap verbs, expand builds. Six weapons, move plus collect, shop choices — depth from combinations, not inputs.
- Keep the install tiny. 200 MB plus OpenGL 3+ unlocked mobile and Switch without a rewrite; that is cheaper than any feature.
Do not romanticize:
- Godot does not give you console certification, first-party saves or Game Pass deals for free. Blobfish needed Erabit for mobile and later Evil Empire oversight plus platform holders for PS/Xbox and Switch ports.
- GDScript dynamic typing and signal spaghetti bite at scale — DLC Abyssal Terrors plus 4-player co-op in October 2024 is exactly when architecture debt shows up.
- Survivors-likes now flood Steam. Copying auto-fire plus waves in 2026 without a new constraint just adds to the pile.
Verdict
Brotato did not win because Godot out-renders Unity. It won because a solo dev picked an engine that asked almost nothing — no fees, 28 to 189 MB editor, 2D scenes plus GDScript — and then designed a game that asked almost nothing back: move, survive 90 seconds, shop, repeat in under 30 minutes.
If you are starting a 2D systems game, ask the Brotato question: what engine lets me balance the shop 100 times this month, and what rewrite will I never need because I stayed boring? If you cannot name the rewrite you avoided, you do not have a stack, you have a hope.
What boring constraint would make your game stranger, not just cheaper? I do not have a rigorous answer. I think living with that question for one prototype month is more useful than any benchmark.
Sources and verification date
Core facts verified October 15, 2026 via direct English retrieval over HTTP. Analysis and provocations are my own and labeled as such.
- Engine Godot, developer Thomas Gervraud Blobfish Lyon fourth game, early access September 27 2022 full June 23 2023 mobile March 28 2023 Switch August 3 2023 PS Xbox January 30 2024 DLC Abyssal Terrors co-op October 2024 transfer Evil Empire September 2025, 1M early access overwhelmingly positive most popular roguelite Steam Deck nomination
- Godot 2022 retrospective naming Brotato Dome Keeper Case of Golden Idol as hits, Godot 4.0 alpha beta timeline
- Steam auto-fire manual option fast runs under 30 minutes waves 20 to 90 seconds dozens characters hundreds weapons flamethrowers SMGs local co-op 4 players 2 Ghz 4 GB 128 MB OpenGL 3+ 200 MB
- Godot free open-source MIT C++ C# GDScript nodes scenes signals file system no fees 28 to 189 MB
- Godot free open-source 2D 3D pitch
- Iron Snout endless waves Classic Sudden Death Wolfieball browser fighting
- Stick Merge merge pistols to SMGs assault rifles damage scaling progression
Top comments (0)