NFT Rarity Explained: Traits, Scores, and Collection Value

What Makes an NFT Rare?

Most NFT collections are generative: a program assembles each token by combining a set of visual layers or attributes. A profile-picture collection might layer backgrounds, body types, clothing, accessories, and facial expressions. Each layer is a trait category, and each specific option within that category is a trait value.

Rarity comes from the distribution of those trait values across the collection. If 80% of tokens have a blue background but only 2% have a gold one, a token with a gold background is rarer. The fewer tokens share a particular combination of traits, the rarer that specific token is considered.

Some projects also include a special trait called 1-of-1s — completely unique tokens hand-crafted by the artist rather than generated. These are almost always the most valuable items in a collection because no other token shares their appearance.

Rarity is determined entirely at mint time. Once a collection is deployed, the distribution of traits is fixed. This makes rarity an objective, verifiable property of each token — anyone can check it against the on-chain metadata.

How Rarity Scores Are Calculated

Several mathematical formulas are used to turn trait frequencies into a single comparable number. The most widely used methods are:

Trait Rarity (Simple)

The simplest approach assigns each trait a rarity percentage — the share of the collection that has that trait. A token's overall score is the average rarity of all its traits. This method is intuitive but can produce misleading results because it weights all traits equally regardless of how much visual impact they have.

Rarity Score (OpenRarity / Rarity Tools method)

A more robust method calculates the statistical rarity of each trait using its probability: rarity score for a trait = 1 ÷ (number of tokens with that trait ÷ total supply). Each trait's score is then summed to produce a total rarity score. Higher scores indicate rarer tokens. This is the basis of the formula popularized by the original Rarity Tools website and adopted by many subsequent platforms.

Information-Theoretic Rarity (OpenRarity standard)

OpenRarity, an open standard developed by major marketplaces including OpenSea, uses an information-theoretic approach based on statistical entropy. It is designed to be more mathematically rigorous and resistant to gaming by projects that add meaningless traits to inflate scores. The formula measures how much information each token's trait combination encodes — tokens that are harder to predict from the collection's distribution receive higher scores.

In practice, different tools may rank the same token differently depending on which formula they use. When comparing rarity scores, always confirm which method the tool is applying.

Popular Rarity Ranking Tools

Several tools exist specifically for looking up rarity scores and rankings within collections:

Rarity Sniper

Rarity Sniper is one of the most widely used rarity checkers. It covers thousands of Ethereum and Solana collections, shows per-trait rarity percentages, and ranks each token within its collection. It also provides live floor price data, making it easy to compare a token's rarity rank against current market pricing.

OpenRarity

OpenRarity is an open-source standard and accompanying website maintained by OpenSea, Curio, icy.tools, and PROOF. It applies the information-theoretic scoring formula described above. Because it is open and auditable, it is increasingly adopted as a neutral reference, particularly for high-value Ethereum collections.

HowRare (Solana)

HowRare.is focuses on the Solana ecosystem and is the de facto rarity standard for most Solana collections. It is integrated directly into Magic Eden's marketplace, so rarity ranks often appear alongside listings without needing a separate lookup.

Marketplace-Native Rankings

Major marketplaces increasingly embed rarity data natively. OpenSea displays rarity ranks powered by OpenRarity for supported collections. Magic Eden shows HowRare ranks for Solana NFTs. Getgems, the primary TON NFT marketplace, does not yet have a standardized on-site rarity ranking system, so TON collectors typically assess rarity manually using collection metadata or community tools.

Rarity Across Ethereum, Solana, and TON

The concept of rarity applies universally to generative NFT collections, but the tooling and culture around it differ across blockchains.

Ethereum

Ethereum has the most developed rarity analysis infrastructure. Rarity tools are mature, widely integrated, and supported by a large collector community that actively uses rarity data when making purchasing decisions. Blue-chip collections like CryptoPunks and Bored Ape Yacht Club have well-documented rarity hierarchies that directly map to market prices.

Solana

The Solana NFT ecosystem has strong rarity tooling through HowRare.is and its Magic Eden integration. Solana collectors are generally rarity-conscious, and rare traits in popular collections can command significant premiums despite the overall lower price floor compared to Ethereum.

TON

The TON ecosystem, with its integration into Telegram, is newer to generative collection conventions. Many TON NFTs — including Telegram's collectible gifts and Fragment usernames — are valued based on factors other than generative trait rarity: username desirability, gift edition size, and social signal. For generative collections on Getgems, rarity analysis requires examining on-chain metadata directly, as centralized rarity tools are less established here.

How Rarity Affects Price

Rarity is one of the most reliable predictors of relative price within a collection, but it is not the only one. Several dynamics connect rarity to market value:

The Floor-to-Rare Premium

Every collection has a floor price — the lowest price at which any token is listed for sale. Rare tokens typically trade at a multiple of the floor. Top-ranked tokens in popular collections can sell for 5x, 10x, or even 100x the floor, depending on the collection's cultural status and the perceived uniqueness of the traits involved.

Trait Desirability vs. Statistical Rarity

Statistical rarity and community desirability are not always the same thing. A trait might be mathematically rare but aesthetically unpopular — and the market will price it accordingly. Conversely, some traits are so visually striking or culturally loaded that they command premiums beyond what raw rarity scores would suggest. Gold or laser eyes in ape-style collections, alien body types in profile picture projects, and clean backgrounds in minimalist collections are examples where community preference amplifies or overrides pure rarity math.

Liquidity Considerations

Very rare tokens can be harder to sell quickly. The pool of buyers willing to pay a rare-tier premium is smaller than the pool looking for floor tokens. This creates a liquidity trade-off: rare tokens may hold or grow in value better over time but can sit unlisted longer when you want to sell. Collectors who prioritize flexibility often balance their holdings between some rare pieces and more liquid floor-adjacent tokens.

Limitations of Rarity Rankings

Rarity scores are useful tools, but they have real limitations worth understanding before making decisions based on them.

Scores vary by tool. Different rarity platforms use different formulas, and they can disagree significantly on a token's rank. A token ranked in the top 1% on one platform might rank in the top 5% on another. Always confirm which standard you are comparing against.

Rarity is collection-specific. A rarity rank of 50 out of 10,000 says nothing on its own — it only has meaning relative to that collection's specific distribution and the market demand for that project. A top-ranked token in a forgotten collection may be worth less than a mid-ranked token in a thriving one.

Trait additions can distort scores. Some projects have added traits or revealed hidden attributes after mint, which can change rarity distributions retroactively. The OpenRarity standard was partly designed to reduce the impact of such manipulations, but older tools and manually maintained databases may not reflect updates immediately.

Markets ultimately decide. Rarity rankings are a signal, not a guarantee. A high rarity score does not ensure strong demand or a quick sale at a premium price. Market conditions, collection reputation, cultural momentum, and individual buyer preferences all contribute to actual prices. Treat rarity as one input among several rather than the sole determinant of value.

Conclusion

Rarity is one of the most fundamental concepts in NFT collecting. In generative collections, each token's combination of traits produces a unique statistical fingerprint, and scarcer combinations tend to command higher prices. Rarity scores give collectors a standardized way to compare tokens within a collection and inform buying and selling decisions.

The most useful approach is to check rarity using a trusted tool for your specific blockchain, understand the underlying formula being used, and weigh rarity alongside other factors like collection reputation, trait desirability, and your own liquidity needs. Tools like NFT Bowl let you view your entire multi-chain collection — Ethereum, TON, and Solana — in one place, making it easier to survey what you hold and research which tokens in your portfolio might carry rare traits worth tracking.