Palworld Breeding Calculator

Palworld assigns every Pal species a hidden breeding power value. When two Pals produce an egg, the game computes the child in three steps:

1. Target power = floor((parent 1 power + parent 2 power + 1) ÷ 2)
2. Offspring = the standard Pal whose breeding power is closest to the target (with an internal priority value breaking ties)
3. Overrides — same-species pairs always produce that species, and roughly 160 hard-coded special combinations produce fixed children (this is how unique variants are bred)

Worked example: Lamball has breeding power 3050 and Cattiva 2760. The target is floor((3050 + 2760 + 1) ÷ 2) = 2905, and the Pal with the closest breeding power to 2905 becomes the offspring. Because many Pals cluster near each other in power, small parent changes can produce surprising children — which is exactly why a calculator beats guessing.

Rather than approximating this formula, this calculator embeds the actual breeding table extracted from the game files — every one of the 44,850 unique parent pairs and their exact child, including all special combinations. No approximation, no missed edge cases.

Breeding power — called CombiRank in the game files — is a hidden number the game assigns to every Pal species. It is the single value that drives every breeding result, yet it is never shown anywhere in game, which is the entire reason breeding calculators exist.

In the current game data it ranges from 20 to 3100 across 288 distinct values. The scale runs backwards from what you might expect: lower numbers mean rarer, stronger Pals, and higher numbers mean common early-game ones.

The inverse relationship is consistent rather than anecdotal. Measured across all 299 Pals in the current data, breeding power correlates negatively with base attack (−0.80), defense (−0.76), HP (−0.73) and the game's own rarity value (−0.73). So as a rule of thumb, a Pal with a low breeding power is one you had to work for.

It is an ordering index, not a power level. Breeding power exists to place every Pal on one line so the game can answer "what sits between these two parents?" The correlation with strength is strong but not exact — Panthalus has the lowest breeding power in the game (20) yet Jetragon (70) has higher attack and a higher rarity rating. Never treat it as a tier list.

Why it matters twice. Breeding power feeds both systems on this page. Normal breeding averages the two parents' values and picks the nearest Pal, so two parents on opposite ends of the scale produce a mid-scale child. Mutation instead uses the lower value to size the outcome band and the difference to position it — which is why a rare parent makes mutations predictable, as shown above.

Every egg produced at a Breeding Farm rolls an independent chance of coming out mutated — about 1% per egg with a standard Cake, or about 3% with an Extravagant Vegetable Cake. A mutated egg hatches a different species than the pair's normal child, and that Pal can carry exclusive passive skills that normal breeding and wild capture never produce.

Which species you get is not random across the whole roster — it is determined by the parents. The game takes the lower of the two breeding powers and their difference, then derives a narrow band of candidate ranks:

1. count = max(1, round(low × 0.1)) — how many ranks are in play
2. start = round(low × 0.5) + round(diff × 0.4) + 1 — where the band begins
3. The game rolls one rank in that band and hatches the eligible Pal nearest to it

Instead of simulating a single roll, the Mutations mode enumerates every rank in the band, so the percentages it shows are the exact outcome distribution rather than an estimate. Pick two parents above to see every species their mutated eggs can produce.

Separately, because each egg's mutation roll is independent, the odds never "build up" — 100 eggs at 1% gives roughly a 63% cumulative chance of at least one mutation, not a guarantee. The calculator uses 1 − (1 − p)ⁿ for that, plus thresholds for 50%, 90%, and 99% confidence.

The single most useful thing to understand about Palworld mutations is that the lower-ranked parent controls how predictable the result is. Because count = max(1, round(low × 0.1)), the band of possible ranks is one tenth of the lower breeding power — and rarer Pals have lower breeding powers.

In practice this means breeding with a rare Pal produces a narrow, often single-outcome mutation, while breeding two common early-game Pals scatters the result across dozens of species:

How to target a specific mutant: put one rare, low-breeding-power Pal in the farm to shrink the band, then vary the second parent. Changing the second parent moves start (through the difference in breeding powers) while the band stays narrow — so you can slide the whole window onto the species you want. Checking pairs in the Mutations mode before committing farm time is far faster than testing in game, since roughly 8% of all parent pairs produce a single guaranteed mutation.

Same-species pairs still mutate into something else. Two Lamball always produce a Lamball normally, but a mutated egg from that pair hatches an entirely different species — of all 299 Pals, only one can mutate into itself. This is why players sometimes get an unexpected Pal from a pair they assumed was "locked" to one species.

Not every Pal can appear as a mutation. 39 of the 299 Pals are flagged in the game data as ineligible for rank matching — the tower-boss tier and certain unique variants — so they can never hatch from a mutated egg no matter which parents you use. The calculator only ever lists the 260 Pals that are genuinely reachable.

Use Find Parents mode before hunting: if your target Pal has hundreds of valid parent pairs, you almost certainly already own a workable combination — check what you have before catching anything new. If it has none, you know immediately that wild capture or same-species breeding is the only path, saving hours of trial and error at the breeding farm.

Breeding power determines the species, but passive skills and IVs inherit separately from both parents. Serious breeders first reach the target species, then repeatedly breed within that species (same-species pairs always produce the same species) while selecting parents with the desired passives until the traits stack.

Cakes are consumed per egg produced, so plan batches: work out your full breeding chain here first — including intermediate steps — before committing farm time and resources to a multi-generation plan.