Mechs vs Kaijus
1,000 Mechs against 10 Kaijus — a count ratio of 100 : 1. The official description's own example. A hundred Mechs per Kaiju is the kind of extreme count ratio the sandbox is built to test.
What the game's 100+ creatures are, the four the official description names, the two example matchups with real counts, and how to design a build worth testing.
The official Roblox description advertises 100+ creatures across many categories and publicly names exactly four: Mechs, Kaijus, Tigers, and Elephants. It also publishes two example matchups with exact counts. No creature stats, roles, or a full roster list exist in any official or community source as of 2026-10-03.
Everything below comes straight from the game's own Roblox page. Treat these as confirmed vocabulary — the rest of the roster stays in-game until the creator publishes more.
| Creature | Where it is named | Role in the official examples |
|---|---|---|
| Mechs | official description, 1000-vs-10 example matchup | Swarm-side creature in the game's flagship example |
| Kaijus | official description, 1000-vs-10 example matchup | Small-count counterpart to the Mechs |
| Tigers | official description, 500-vs-200 example matchup | Mid-ratio example side A |
| Elephants | official description, 500-vs-200 example matchup | Mid-ratio example side B |
Source: official Roblox game page description, checked 2026-10-03.
These two matchups come verbatim from the description. Their counts are the anchor values this site's calculators use for worked examples.
1,000 Mechs against 10 Kaijus — a count ratio of 100 : 1. The official description's own example. A hundred Mechs per Kaiju is the kind of extreme count ratio the sandbox is built to test.
500 Tigers against 200 Elephants — a count ratio of 2.5 : 1. The official description's second example. A closer 2.5-to-1 ratio, showing the game supports anything from fair fights to absurd ones.
Source: official Roblox game page description, checked 2026-10-03.
Progression games define a build as equipment plus stats. This game has no equipment, so a build is the battle you compose — and the difference changes how you should think about it.
Every battle is a comparison: a count of one creature against a count of another. The official examples bracket the useful range — 1000 vs 10 answers questions like can a swarm overwhelm a giant, while 500 vs 200 asks whether quality beats numbers at a 2.5-to-1 handicap. Start from a question you actually want answered, then set the counts that test it.
If you swap creatures and counts and map at once, you cannot say what caused the result. Run the same matchup once to see the baseline, then adjust a single number — 500 Tigers become 600, or the Elephants stay put — and rerun. This is the cheapest way to find the crossover point where the winner flips, and it is exactly what the site's matchup and army-size planners are built to record.
The description promises realistic physics and total chaos, which means bodies interact in ways that are not fully deterministic. A landslide result is trustworthy; a near-draw deserves two or three replays before you conclude anything. Writing down the result of each replay — the squad planner does this — turns luck into a trend line.
Bigger battles are the game's headline promise (like-unlocked, per the official description), but scale is a variable, not a goal. A 20-vs-20 probe tells you more per minute than a maximum-scale brawl whose result you cannot read. Save the giant battles for confirming what small battles suggested.
Recorded as absent as of 2026-10-03 — this list shrinks only when the creator publishes something new.
The description says 100+ creatures but names only Mechs, Kaijus, Tigers and Elephants. No complete roster or category list has been published.
No official or community stat table exists. This is a sandbox, not a progression RPG — creatures have no published numbers to rank.
Only different maps is documented. Names, sizes and per-map battle limits are unpublished.
The like-unlock system is official, but no like counts or battle-size thresholds have been announced.
No public code tracker covers this game and no code system is documented. An empty code list here is a research finding, not an omission.
The developer has not published a patch-note channel. Roblox's API records only a last-updated timestamp.
The official Roblox description advertises 100+ creatures from every category imaginable, but publicly names only four: Mechs, Kaijus, Tigers, and Elephants. The full roster is visible inside the game and has no published list as of 2026-10-03.
They are the two example matchups from the official description: 1000 Mechs against 10 Kaijus (a 100-to-1 count ratio) and 500 Tigers against 200 Elephants (2.5 to 1). They show the count range the sandbox supports, from absurdly one-sided to nearly even.
No stat table exists in any official or community source as of 2026-10-03. This is a watch-it-happen physics sandbox, not a stats-driven RPG — creature strength is something you test by running the battle.
A good build is a well-designed question: two sides, explicit counts, one map, and one difference at a time. Change one variable per battle and you learn which creature actually wins; change everything at once and the physics chaos gives you an unrepeatable highlight reel.
Each of these uses the officially documented facts as its input.