Guides · Updated 2026-11-02

How to read a Fragment sale history correctly

A sale history is a list of prices, not a price. Read it in four steps: check the date, check what sold (model, backdrop, number), check whether the buyer and seller were independent, and subtract the 5% fee before you compare the result with a listing. Skip a step and the history will tell you a comfortable story that is not true.

Four things every row must tell you

DetailWhy it mattersWhat to do when it is missing
DateA sale from last year describes a different market with different buyersUse the last 12 months as the window
What soldModel and backdrop move the price several times inside one collectionMatch the model first, then the backdrop
IndependenceOne handle resold in 60 days, or a batch at one price, is not a new datapointDrop the row; it is the same money twice
Sale typeBuy-now, auction and timed sale are three different mechanismsCompare like with like, never auction bids with floors

These are the same filters the valuation method applies before a sale is allowed to count.

The traps that make a history look better than it is

The first trap is the whale. One handle sold for 115,617 GRAM on 2026-11-01; that single row says nothing about what an ordinary name is worth, and it does not move the class median because the method drops prices more than 8 times above or below it. The second trap is the round trip: the same handle sold twice in a month is one piece of evidence, not two — usually a flipper testing the market. The third is the batch: ten sales at exactly the same price on the same day are one seller, not ten buyers. The fourth is the current bid: an auction ending at 2,000 GRAM is a bid, and the price is whatever the last bid was when the clock ran out.

There is also a quieter trap: comparing a sale of a rare model with the floor of the whole collection. Snoop Ring, for example, had a model floor of 25 GRAM against a collection floor of 19 GRAM on 2026-11-02. A sale at 25 was a normal sale, not a bargain, and a floor of 19 was never available for that model.

Turn a sale into a net number

Fragment takes 5% of the sale price, so the seller receives 95%. A 100 GRAM sale is 95 GRAM before network and conversion fees, which the site does not include in any figure. Compare that net number with what the item costs you today, not with the headline price, and you will stop calling a 5% loss a flat trade.

Finally, write down the date you read the history. A sale history is a photograph of a market at a moment; the market moves, and the page you are reading may already be a day old.

And count the sales behind any median you are quoted. A median of 6,049 GRAM from 36 sales is a description of a class; a median from two sales is an anecdote with a decimal point. The site's own rule — at least 8 sales after cleaning before a class median is used — exists for exactly this reason, and it is the first thing to check when a valuation looks too confident for the evidence behind it.

Sources: fragment.com public pages (listings, sold items, item pages), the track record page; the site's own open data at /data/floors-daily.json and /data/usernames-daily.json (CC BY 4.0); the valuation method on how we value (as of 2026-10-28). Figures checked 2026-11-02 UTC.

Next: the track record · gift price guide · live floors · the full method

Gift price checkUsername value checkTelegram bot: TG Gifts

Frequently asked questions

How far back should I look at sales?

Twelve months is the window the valuation method uses. Older sales describe a different market, and recent ones carry more weight: a sale's weight halves every 60 days.

Why do you ignore some sales?

Because they are not independent evidence: the same handle resold within 60 days, a batch of identical prices from one seller, prices more than 8 times away from the class median, or sales the TON blockchain contradicts.

Is the highest sale a good reference?

No. A single record sale is an outlier, and one buyer's decision does not describe a class. Use the median of comparable sales and check how many sales are behind it.

Is a sale price the same as what the seller received?

No. Fragment takes 5%, so the seller receives 95% of the sale price before network and conversion fees.

More guides

What Are Telegram Gifts? NFT Gifts Explained (2026)
Telegram Gift Price: How to Value Collectible Gifts (Model, Backdrop, Symbol)
Telegram Gift Rarity: Is My Gift Rare? (Model, Backdrop, Symbol)
What Is a Telegram Gift Floor Price? (Live Floors Explained)
Most Expensive Telegram Gifts Right Now (Live Ranking by Floor)
How to Upgrade a Telegram Gift: Stars Cost and Is It Worth It
How to Sell a Telegram Gift (Collectible NFT Gifts, 2026 Guide)
New Limited Telegram Gift Drops: How to Catch Them First
How Much Is My Telegram Username Worth? (2026 Guide)
Telegram Username Auctions on Fragment: How to Buy a Username
How to Sell a Telegram Username on Fragment (Step by Step)
GRAM vs TON: Why Telegram Gift Prices Are Now in GRAM
Telegram Gifts Glossary: Floor, Model, Backdrop, Symbol and More
What Is TON DNS? How to Buy a .ton Domain (2026 Guide)
What Is Fragment? Telegram's Marketplace for Usernames and Gifts
How Telegram Gift Floors Are Built: Collection, Model, Backdrop
Telegram Username Prices: Length, Words and Patterns (Real Medians)
Telegram Gift Value Ranges by Collection Tier (2026 Data)
How to Verify a Below-Floor Gift Alert in 60 Seconds
Fees and Break-Even on Telegram Gifts: The 5% and What It Hides
Why the Same Telegram Gift Has Different Prices in Different Markets
Timed Sales vs Buy-Now: What Changed on Telegram Gifts
Craft and Upgrade Economics: What You Actually Pay
Why a Rare Trait Can Be Worth Less Than the Floor
Telegram Gift Scams: Seven Patterns and the Check That Stops Each
What Fair Price Means on TG Gifts: Method and Limits