ZVERIKI.COM is a private project built for domestic and wild animals: the ones we feed and take care of.
The project is dedicated to the untimely departed cats Tyson and Boniface, the squirrel Alla Borisovna, her boyfriend Elvis and her mother — Evolution.
Legal matters: legal@zveriki.com
Partnerships and joint projects: partnership@zveriki.com
Advertising: adv@zveriki.com
Other questions: feedback@zveriki.com
Images, videos and other content created or uploaded by users belong to them or to other rights-holders. The project does not acquire exclusive rights to these materials.
If a user has opened a profile and shared materials with the project, they are published under CC BY-NC 4.0 or under the ZVERIKI licence — as specified in the terms of use.
The project materials were prepared using ZVERIKI's own data, the data listed in the dataset at https://zveriki.com/dataset/dataset.tsv, and open scientific, cartographic and other sources:
ZVERIKI's own data, user data and the listed scientific sources were additionally processed using ML and AI methods.
Data in the project may be inaccurate owing to contradictions in the source materials and to the extrapolation procedure.
The calculators rely on species-average data for the life stage and sex. Individual characteristics of a particular animal — activity, diet, past illnesses, breed and personal measurements — are not yet included in the calculation; the project is working on this.
One of ZVERIKI's goals is to learn to forecast an animal's lifespan from data on its activity, diet and illnesses. For now the tools are neither veterinary nor dietetic. The project wants to collect such data and share it with the world for science and for more confident predictions.
Calculator results are a vivid analogy, not a zoological, veterinary, dietetic or psychological conclusion.
AgeCurculator, SizeCurculator, MassCurculator, EatCurculator and SgrabCurculator are an intentional pun: calculator and curculator (from Latin circus — circle). Age, size, mass, energy and the hug ring all form a comparison between a person or a pet and another living being.
In all the calculators the target organism is the person or pet you start from. The model organism is the animal you compare with. The two sides can often be swapped.
Translates a person's age into the age of any other animal. Not "multiply by seven". Animals age differently: some have a childhood of one month, others a decade; some live to 5, others to 200. AgeCurculator builds an individual conversion curve for each "person → animal" pair, based on real biological data.
The conversion is based on interpolation between life-cycle checkpoints: birth, sexual maturity, adulthood, average and maximum lifespan. Events shared by both species are aligned to draw the curve. If a person is now X years old, the curve says how many years of the chosen animal's life that corresponds to.
Example. A person is 40 years old. The mallard (Anas platyrhynchos) lives on average 5–10 years, up to 29 at most. The checkpoints are aligned: human sexual maturity ≈ mallard sexual maturity, human average lifespan ≈ mallard average lifespan. The curve translates 40 human years into ~8 mallard years.
A date of birth is often known only approximately. The calculator accepts a date range — from a precise day to a month or year — and returns an age with uncertainty. The more precise the date, the more precise the translation.
Animals do not have a year of 365 days. Every species has its own period: the number of days that counts as one "year" in that animal's world. The period shows how often a "birthday" of that species occurs in the human calendar: from several times a day for short-lived insects to once a century for very long-lived fish. AgeCurculator is not limited to one pair. For every birthday of the animal it calculates in advance how old the person will be at that moment. The result is stored as a chain of birthdays and anniversaries and feeds the upcoming-events feed.
That is why a capybara celebrates its birthday several times a year: its life is shorter than a human's, its "year" lasts about half a year in our calendar — and a birthday in the capybara's world comes round more often than once per human year.
The calculator computes the chain of birthdays and anniversaries in advance and assembles the near-future events feed. The default model is the animal on the open profile; it can be replaced via search if life-cycle data are available. If males and females have different lifespans, a separate curve is built for each sex.
In short: date-of-birth range → model age with uncertainty → next birthday and anniversary → events feed.
Compares the linear dimensions of a person or pet with those of any animal. Not "roughly like a cat", but an exact ratio.
Three types of dimension are compared: length — body length, height — height at the shoulder or standing height, circ — circumference. Circumference does not stand in for length or height. If one organism has a length and the other a height, comparison is still possible: the tool shows that different types are being compared and notes this.
The target dimension is entered in millimetres or taken from the pet profile. For the model animal the life stage and sex are chosen, where data differ by sex. By default the adult stage is used — or the stage closest to the target organism's age, if the age is known.
The result is the model's dimension, the "target is larger than model" ratio and its inverse. Units adjust to the magnitude: millimetres, centimetres or metres.
In short: dimension type → model stage and sex → ratio in both directions.
Compares the mass of a person or pet with the mass of any animal. The target mass is entered manually or taken from the profile.
Mass depends heavily on age. The stage and sex are chosen for the model. If the target organism's age is known, the stage is selected using the same age-conversion logic as in AgeCurculator: the equivalent age is translated into the model's age, and the nearest stage is taken from that.
Within a stage, mass is not a point but an interval. For taxa and breeds (not for a specific pet) the calculator shows the interval and lets you pin a point inside it. If both mass and body size are available, refining the size extrapolates the mass: a larger individual is predictably heavier.
The result is the model's mass and two ratios. Units are grams or kilograms.
In short: target mass + age → model stage and sex → ratio in both directions.
Answers the question: if the target organism "ate" the model animal, how long would it last. This is not a menu — it is a comparison of the target's daily energy expenditure with the energy stored in the model's body.
The daily requirement is calculated by Kleiber's law: proportional to body mass to the power of 0.75. The target's age refines the model stage: a juvenile and an adult hold different calorie reserves.
Not all of the body is assimilated. Bones, fur, feathers and chitin are counted as ballast. The inedible fraction depends on the taxon: one ratio for a bird, another for a fish, another for an insect.
The result is how long one individual lasts and how many individuals are needed per day.
If the model is a carnivore or omnivore, the comparison can be reversed: see how long the target would last for the model.
In short: target mass → Kleiber requirement → model mass and stage → calories minus inedible fractions → duration and individuals per day.
It asks whether an animal fits in one person’s hug — and if not, how many people are needed to hug it together. A squirrel or a kitten can be gathered into a palm. An elephant cannot: a chain of people is already needed around it.
How the word is built. The existing term girth names a complete encirclement: a tape around the widest place of an animal standing open.
English has no everyday measuring noun for a sharp, incomplete, messy grab. By analogy with *girth* the project coins grabth — from the verb *grab* and the measuring ending of *girth*.
Grabth is a quantity that characterises whether an animal can be seized with the hands (or other limbs) without fully closing around the trunk. In practice this is the length of the ring around the animal in the pose in which it is naturally picked up: a ball, tucked legs, folded limbs, a coiled snake. It is not girth.
The target organism is the one who hugs. If no measure is given, an adult-human guide is used: a useful hug girth of about 1.6–2.0 m. The model organism is the animal from the profile. Stage is chosen with the same age logic as in AgeCurculator and MassCurculator. The same ring can be read as a hatch, a carrier or a burrow.
The grabth shape is the most compact pose that does not break the skeleton. The calculation uses spine flexibility, the longest barely bendable bone, trunk girth and the contribution of tucked limbs. A species has folding coefficients for length and width: a python folds a lot, an elephant almost not at all.
Grabth is divided by one person’s hug. If the result is one — it is a hug. If more — a circle of people. For a person with a hug of about 1.8 m the typical order is: cat and python — one, giraffe — three, elephant — five, blue whale — eight.
SgrabCurculator uses species folding coefficients and mean stage measures, not the flexibility of a particular individual. Fur, horns, tusks and a spread tail fluke can enlarge the ring beyond the table value.
The tool does not model violence to the skeleton. It is not a zootechnical permit for transport and not an instruction to hug wild animals. Hugs in the concept are a household size scale, not advice to approach an elephant or a whale. For a cage, a carrier or a burrow this is an order-of-magnitude estimate, not a certificate.
When the profile holds real pet measures — chest girth, body length, thigh length — the project aims to put them into the calculation.
In short: girth, bone and folding coefficients → grabth → how many people close the ring.
IPIP+22 is a 22-axis bipolar model for a situational test and animal-archetype matching. The name is built from International Personality Item Pool and the number of axes. It is an applied model of the project: it matches answers against the animal database, it does not diagnose and does not describe a person's character.
An archetype here is not a personality type. It is an animal in a specific cultural tradition, plus the qualities that tradition attributed to it. A real, non-extinct animal is taken, the culture or image source is noted, a species linked to that territory is chosen, and the qualities are mapped onto the axes. The same species in different cultures gives different archetypes.
Each axis is bipolar, with a working range of −10 to +10. Zero means the trait is not expressed or there is too little data. The axes: integrity, sensitivity, sociability, agreeableness, conscientiousness, curiosity, intellect, spirituality, aesthetics, resilience, sacredness, harmony, corporeality, eternity, physical strength, autonomy, attractiveness, status-seeking, playfulness, traditionalism, rebelliousness, self-sacrifice.
A test item is a situation and a set of reactions, from a strong low pole through the midpoint to a strong high pole. There is always a "don't know" option: choosing it does not shift the profile. A choice may slightly touch adjacent axes: the reaction is read as a gesture in a situation, not a direct admission of a trait.
For each axis a value and a confidence are computed. Confidence drops if there are few answers, "don't know" was chosen often or answers on the axis conflict. Weakly filled axes barely contribute to matching. If the picture is still blurry, the test can offer up to three more situations for the same axis; items already shown in the session are not repeated.
First the mode is chosen: for oneself or for an acquaintance, masculine or feminine phrasing. Axes appear in random order. Then close archetypes and complete opposites are built. Unpleasant images are removed; if the list is still long, close relatives of marked ones are removed too. Liked ones gain an advantage in ranking; the profile is not rebuilt. If several candidates remain, the person picks the most similar one. The result shows the chosen animal and its opposites.
The model does not serve for behaviour prediction, "good / bad" evaluation or medical, HR or pedagogical assessment. Axes partly overlap; for comparing cultural images of animals this is acceptable. An archetype describes the traditional image of the animal, not the biology of the species and not a specific person.
In short: situations and reactions → value and confidence per axis → close animals and complete opposites.
The project is in public beta testing. If you notice problems with data or display, please report them via the menu item "🪲 I see a bug".