Why One CS2 Skin Costs Different Money on Nine Marketplaces
A 44-item measurement on live data: Steam runs 47% above the cheapest alternative, a swap bot asks 1.57x the market, and the median spread within a single skin is 1.69x. What the gap is actually made of.
The same item, with the same name, the same wear and the same pattern, costs different money on different marketplaces — and that difference is neither random nor small. We took 44 CS2 items and pulled prices for each one from every marketplace we collect. The median spread inside a single item came out at 1.69x, and the worst case reached 4.9x. Here is what that gap is made of, which part of it you actually keep, and which part only looks like a bargain.
What the measurement showed
The data was taken on 28 August 2026 from live catalogue pages. For every item we took the median price across all trading marketplaces, then divided each marketplace's price by that median. One means "exactly at market", above one means more expensive, below one means cheaper.
| Marketplace | Items measured | Price vs market median |
|---|---|---|
| Steam | 36 | 1.36 |
| lootfarm (swap bot) | 22 | 1.57 |
| CS.Deals | 6 | 1.14 |
| Skinport | 35 | 1.03 |
| Market.CSGO | 38 | 1.01 |
| Waxpeer | 38 | 1.00 |
| Skinout | 6 | 0.99 |
| White.Market | 34 | 0.98 |
| Lis-Skins | 37 | 0.96 |
We also compared Steam against the cheapest alternative for the same item rather than against the market as a whole. That gap is wider: a median of 1.47, meaning buying on Steam costs roughly 47% more than buying on the most favourable third-party marketplace.
Median rather than average, deliberately. An average breaks on a single junk listing: our database has contained prices like $485,937 for a crate genuinely worth five cents. An average would be dragged away by that one row; a median simply ignores it.
Reason one: the Steam fee lives inside the price the buyer sees
Steam charges 15% of the seller's amount — 5% to Steam and 10% to the game's developer. The mechanics matter: the price on the listing is what the buyer pays, and the seller receives that number divided by 1.15.
A seller who needs $20 in hand has to list at roughly $23. The buyer sees $23, and that is the number that lands in our table as "the Steam price". On a third-party marketplace, with a 5-7% fee, the same seller lists at $21.3 and still nets $20. The difference in the sticker price appeared not because a marketplace is greedy, but because the Steam fee is baked into the visible number.
Reason two: Steam money cannot leave Steam
Proceeds from the Steam Community Market land in your Steam wallet and can only be spent inside Steam. There is no official way to cash them out.
That makes a dollar in a Steam wallet not equal to a dollar on a card, and the market prices that in. People who need real money sell cheaper on third-party marketplaces; people who already hold a Steam balance with nothing else to spend it on are happy to pay more. Hence a persistent Steam premium, which will not disappear while the wallet stays one-way.
The practical rule that follows: comparing a Steam price to a third-party price head-on is not a fair comparison. These are prices in two different currencies — locked money and free money. Our table shows both, but only you can decide what a locked dollar is worth to you.
Reason three: a swap bot charges for immediacy
Swap sites (loot.farm, cs.trade, tradeit and similar) work differently from marketplaces: there is no seller and buyer, there is a bot with its own inventory. You hand over a skin and instantly receive another skin or a balance, with no waiting for a buyer.
In our measurement loot.farm asks 1.57x the market median. That is not a fee for nothing: you are paying for the trade happening now rather than in a week. But it should be understood honestly — the bot earns on both sides of its rate. Hand it a skin and you receive noticeably less than market; take a skin from it and you pay noticeably more.
| What you are doing | Marketplace | Swap bot |
|---|---|---|
| Speed of the trade | hours to weeks | seconds |
| Purchase price | close to market | above market |
| What you get for your skin | market minus fee | noticeably below market |
| Who is on the other side | another person | the site's bot |
| Risk of not selling | yes | none |
Reason four: illiquid items drift further apart
The more rarely an item trades, the wider the spread. On liquid items marketplaces converge quickly: a seller who lists well above market simply will not sell. Illiquid items face no such pressure — a listing can sit for months at any price and nobody corrects it.
In the same measurement we looked at real trades: of 60 items checked, 39 had low liquidity and only one was high. The median came to 42 sales in 30 days, about one and a half per day. That is exactly the zone where the wildest price disagreements live.
Reason five: part of the gap is not a price at all
Some of what looks like spread is not a price. Marketplaces carry listings set at impossible numbers: the seller does not want to sell the item, but does not want to remove it either, so they put a four-figure price on it. We have found a crate genuinely worth five cents listed at $485,937, and a key at $497,005.
Such rows have nothing to do with the market, and we filter them out: a price more than 20 times above the reference never enters our database. But if you are comparing marketplaces by hand, remember that the top number in the list is very often not "expensive" — it is a placeholder listing.
- Look at the range across all marketplaces, not at a single price — one number tells you nothing.
- Judge Steam separately: different currency (locked money) and the fee sits inside the sticker.
- Discard extreme highs — they are placeholder listings more often than offers.
- When selling, count what reaches you after fees, not the listing price.
- If the item is illiquid, budget for weeks of waiting or accept a discount for speed.
How it looks on one specific item
Take AK-47 | Cartel (Field-Tested) on the day of the measurement. Skinport — $14.10, Lis-Skins — $14.56, Skinout — $14.65, White.Market — $14.95, Waxpeer — $15.03, Steam — around $21. The spread between the extreme third-party marketplaces is under 7%, while Steam stands apart, and it is Steam that creates the impression of a huge difference.
The conclusion is straightforward: third-party marketplaces compete with each other and stay close together, and the gap up to Steam is a structural premium rather than somebody's pricing mistake. A genuine mispricing is what you find when a single marketplace breaks away from the tight cluster of the others — that is either a stale quote or a genuinely underpriced listing.
Every item page in our catalogue shows the range across all marketplaces at once, plus a separate "what you actually get paid" block — net proceeds after the specific marketplace's fee. That removes the manual arithmetic the whole table was built to replace.
Caveats about the method
We deliberately quote the numbers of this measurement rather than "market averages" in general. The sample is 44 CS2 items, mostly liquid rifles plus a few capsules; it contains no expensive knives or gloves, where spreads run wider. Prices were captured on a single day, and they move. Marketplaces appear in different numbers of items: Waxpeer and Market.CSGO were present on 38 cards, while CS.Deals and Skinout only on six, so their coefficients are less stable.
What follows from that: the order of magnitude is reliable (Steam is dearer, swap bots are dearer, P2P marketplaces cluster tightly), while the exact percentages should be read as one day's snapshot rather than a constant.
Price spread is not chaos but a sum of understandable causes: a fee inside the sticker, locked money, a charge for immediacy, the liquidity of the specific item, and junk listings. Break the difference into those parts and it becomes clear which portion is a real chance to buy cheaper, and which is an illusion you will pay for at the other end of the trade.