Counted from published rosters
Two studios appear on all five rosters read here. Seventy-nine appear on one
Thirty-three of the hundred operators in this library publish a studio roster that can be counted. Between them those rosters name 454 distinct studios, and not a single studio appears on all 33 — which is the first sign that a game library is a negotiated list rather than a catalogue.
What is a game library actually made of?
Not games. Studios. A crypto casino does not build the titles it shows; it takes feeds from software suppliers, and the set of suppliers it has signed is the real boundary of what the site can ever display. Everything else — the search box, the categories, the count in the header — sits on top of that set.
So the countable object is the roster, and rosters are unevenly published. Of the hundred operators read for this library, 33 publish one that can be counted from their own provider index, filter or sitemap. The other 67 show logos and leave the number unstated, which is the same move the coin strip makes on the deposit page and is dealt with on the page about coin lists.
The 33 rosters run from 12 studios to 150. The middle one holds 47.
Five of the ten operators compared on this site are among the 33: Bitcasino.io at 120 studios, Empire.io at 79, CoinCasino at 63, Bitsler at 41 and BetPlays at 19. The other five publish nothing countable, and a blank there means the list was looked for and not found rather than that the operator has no suppliers.
How much do two big libraries have in common?
Less than their sizes imply. Put the 33 rosters together and they name 454 distinct studios. Of those 454, 180 appear on exactly one roster in the whole library, and none appears on all 33. The most widely carried names — Endorphina and Pragmatic Play — reach 26 rosters of the 33.
Narrow it to the five rosters read at the operators on this site and the picture sharpens. Those five name 172 distinct studios between them. Seventy-nine of the 172 appear on exactly one of the five. Two appear on all five: Betsoft and Hacksaw Gaming.
Two studios out of 172 are what these five sites reliably share.
The same holds at the top of the market. The two longest rosters anywhere in this library are Instant Casino's 150 and Betriot's 140; they have 54 studios in common out of 236 between them, so under a quarter of the combined set is on both. Two libraries can each be enormous and still overlap thinly, because each was assembled from separate commercial agreements rather than from a common catalogue.
What decides whether a title opens for a given player?
The roster is a maximum, not a promise. Between the roster and the screen sit two restriction lists that do different jobs, and the terms of five operators in this library keep them explicitly apart.
The first is the account-level restricted-country list, which decides whether the account may exist and play at all. Thirty-eight of the hundred publish one. They run from a single named country to 224 entries, and the middle list names 34.
The second list belongs to the studio, not to the casino.
Bitdreams' terms carry game-provider restrictions naming NetEnt and NYX that are, in the operator's own framing, separate from the account-level list. PlayAmo does the same, naming NetEnt and NYX/NextGen. Oshi Casino attaches restrictions to NetEnt titles further down its terms. Bitkingz goes furthest and names the titles: NetEnt games, Street Fighter, Vikings, Universal Monsters and its jackpot games. CryptoLeo puts game-specific availability restrictions in a section of its terms that is not the restricted-countries section at all.
Five operators, five documents, one structure. The casino signs a supplier; the supplier arrives with its own licensing map, usually inherited from the studio's own regulators and from the intellectual property behind branded titles; and that map is enforced by the studio or its aggregator before the casino is involved. A title can therefore be on the roster, visible in the lobby, and refuse to load, without anything about the account having changed.
Why the number in the header is never the number you get
Because four separate filters stand between the roster and a launched game, and only the first is advertised. The supplier has to be signed. The supplier has to permit the territory. The operator's own country list has to permit the account. The licence the operator holds has to permit the market — and on this site's set, three of the ten hold licences whose standard conditions name jurisdictions their own country lists do not.
None of those four is a number an operator prints beside its game count.
Which is the practical reading of any headline library figure: it describes the catalogue the operator has bought, not the catalogue a given account can open. The count is real, it is just answering a different question than the one it appears to answer.
What this article does not claim
No game was launched, no lobby was loaded and no availability was tested from anywhere. Every figure above is counted from rosters and restriction lists published by the operators themselves.
The rosters carry a further limit worth stating plainly. Several were read from archived copies of the operators' own provider pages — Bitcasino.io's 120, for instance, comes from snapshots taken between 6 February 2023 and 18 July 2026 — so each is the roster as of its snapshot rather than as of today. That distinction, and what an empty cell is allowed to mean, is set out in how the figures are read. What Bitcasino.io publishes elsewhere in its terms is set out on its own page.