Here crafts are calculated at Steam prices: with fees, over all possible outcomes and with the needed float for every input.
Craft — ten items of one rarity are exchanged for one item of the next rarity from the same collections. Which one drops is decided by chance, but the odds and the float of the outcome are computed exactly.
Invest — how much the inputs cost at the cheapest lots. Returns — the average over all outcomes after the Steam 15% fee. Profit — their difference, on average over many crafts. Chance of profit — the share of outcomes that pay for the input on their own.
Buy float — the strictest wear requirement among the inputs: lots with a float above this number don’t fit. “Any lot” — no need to choose.
Before buying, open the craft and press «Check floats on Steam»: the site will look at real listings and recalculate the profit.
Crafts calculated in total: 43 041.
Profitable33craftsof 34 723 in the recalculation
Best profit—nothing profitable to assemble
Median ROI+2%half of the profitable crafts are above
The summary covers all crafts of the recalculation, risk-labelled ones included: 34 723.Of them the site shows the best: 15 000.Shown in the list below: 546.Hidden:“profit only from resale” — 248,
“outcome above the Steam cap” — 3 613.
A craft can carry several labels.Your own conditions apply too.
Craft ROI all crafts of the recalculation: how many in each band
No profitable craft that can be assembled under these conditions: not enough market lots, or the profit rests on resale and no-deal prices. The full list is below.
FAMAS | Rapid Eye Movement (Field-Tested): buy 0.1500–0.1829 (≈2% of this wear's listings); Sawed-Off | Devourer (Field-Tested): buy 0.1500–0.1829 (≈2% of this wear's listings)
invest at least 15 356 ₽with floatassemble: ≈ won't work
float premium: at least 9 031 ₽
profit on prices without salesat least −5 941 ₽rare inputsThe rarest input sells 0 times a day. Buying them up will take long, and the price will change meanwhile.
invest at least 15 768 ₽with floatassemble: ≈ won't work
float premium: at least 9 381 ₽
profit on prices without salesat least −6 133 ₽rare inputsThe rarest input sells 0 times a day. Buying them up will take long, and the price will change meanwhile.
invest at least 3 437 ₽with floatassemble: ≈ won't work
float premium: at least 2 016 ₽
profit on prices without salesat least −476 ₽rare inputsThe rarest input sells 7 times a day. Buying them up will take long, and the price will change meanwhile.
StatTrak™ UMP-45 | Grand Prix (Field-Tested): buy 0.2500–0.2551 (3% of this wear's listings); StatTrak™ CZ75-Auto | Eco (Field-Tested): buy 0.1500–0.1618 (under 1% of this wear's listings)
Stays in profit even if the week before the sale turns out bad: only one week in ten is worse.
invest at least 3 420 ₽with floatassemble: ≈ won't work
float premium: at least 1 995 ₽
profit on prices without salesat least −156 ₽rare inputsThe rarest input sells 12 times a day. Buying them up will take long, and the price will change meanwhile.
StatTrak™ UMP-45 | Grand Prix (Field-Tested): buy 0.2500–0.2551 (3% of this wear's listings); StatTrak™ MP7 | Powercore (Field-Tested): buy 0.1500–0.1618 (under 1% of this wear's listings)
Stays in profit even if the week before the sale turns out bad: only one week in ten is worse.
MAG-7 | Petroglyph (Factory New): buy 0.0000–0.0037 (≈1% of this wear's listings); Nova | Toy Soldier (Field-Tested): buy 0.1500–0.1622 (under 1% of this wear's listings)
Stays in profit even if the week before the sale turns out bad: only one week in ten is worse.
UMP-45 | Grand Prix (Field-Tested): buy 0.2500–0.2551 (under 1% of this wear's listings); Nova | Toy Soldier (Field-Tested): buy 0.1500–0.1618 (under 1% of this wear's listings)
Stays in profit even if the week before the sale turns out bad: only one week in ten is worse.
UMP-45 | Grand Prix (Field-Tested): buy 0.2500–0.2551 (under 1% of this wear's listings); MP7 | Powercore (Field-Tested): buy 0.1500–0.1618 (under 1% of this wear's listings)
Stays in profit even if the week before the sale turns out bad: only one week in ten is worse.
Negev | Loudmouth (Field-Tested): buy 0.1500–0.1637 (under 1% of this wear's listings); Nova | Toy Soldier (Field-Tested): buy 0.1500–0.1637 (under 1% of this wear's listings)
Stays in profit even if the week before the sale turns out bad: only one week in ten is worse.
MP5-SD | Kitbash (Factory New): buy 0.0000–0.0048 (≈1% of this wear's listings); Nova | Toy Soldier (Field-Tested): buy 0.1500–0.1659 (under 1% of this wear's listings)
Stays in profit even if the week before the sale turns out bad: only one week in ten is worse.
invest at least 4 227 ₽with floatassemble: ≈ won't work
float premium: at least 2 350 ₽
marketplace pricesells slowly62% of the probability falls on outcomes that sell fewer than 21 times a week (3 a day): at the lowest price they won't be bought right away.rare inputsThe rarest input sells 2 times a day. Buying them up will take long, and the price will change meanwhile.
invest at least 3 575 ₽with floatassemble: ≈ won't work
float premium: at least 2 116 ₽
profit on prices without salesat least −117 ₽rare inputsThe rarest input sells 18 times a day. Buying them up will take long, and the price will change meanwhile.
MP7 | Powercore (Field-Tested): buy 0.1500–0.1653 (under 1% of this wear's listings); Nova | Toy Soldier (Field-Tested): buy 0.1500–0.1653 (under 1% of this wear's listings); UMP-45 | Exposure (Factory New): buy 0.0000–0.0046 (≈1% of this wear's listings)
Stays in profit even if the week before the sale turns out bad: only one week in ten is worse.
Dual Berettas | Shred (Factory New): buy 0.0000–0.0075 (≈2% of this wear's listings); XM1014 | Run Run Run (Factory New): buy 0.0000–0.0075 (≈2% of this wear's listings)
Stays in profit even if the week before the sale turns out bad: only one week in ten is worse.
invest at least 3 980 ₽with floatassemble: ≈ won't work
float premium: at least 2 322 ₽
profit on prices without salesat least −380 ₽rare inputsThe rarest input sells 0 times a day. Buying them up will take long, and the price will change meanwhile.
M249 | Aztec (Factory New): buy 0.0000–0.0047 (≈1% of this wear's listings); Nova | Toy Soldier (Field-Tested): buy 0.1500–0.1655 (under 1% of this wear's listings)
P2000 | Amber Fade (Factory New): buy 0.0000–0.0035 (≈1% of this wear's listings); Nova | Toy Soldier (Field-Tested): buy 0.1500–0.1616 (under 1% of this wear's listings)
P2000 | Grip Tape (Field-Tested): buy 0.1500–0.1621 (under 1% of this wear's listings); AUG | Creep (Field-Tested): buy 0.1500–0.1621 (under 1% of this wear's listings); AUG | Amber Slipstream (Minimal Wear): buy 0.0700–0.0742 (under 1% of this wear's listings)
Galil AR | Black Sand (Field-Tested): buy 0.1500–0.1630 (under 1% of this wear's listings); Dual Berettas | Shred (Factory New): buy 0.0000–0.0039 (≈1% of this wear's listings)
profit on prices without sales at least −5 941 ₽rare inputsThe rarest input sells 0 times a day. Buying them up will take long, and the price will change meanwhile.
profit on prices without sales at least −6 133 ₽rare inputsThe rarest input sells 0 times a day. Buying them up will take long, and the price will change meanwhile.
profit on prices without sales at least −476 ₽rare inputsThe rarest input sells 7 times a day. Buying them up will take long, and the price will change meanwhile.
profit on prices without sales at least −156 ₽rare inputsThe rarest input sells 12 times a day. Buying them up will take long, and the price will change meanwhile.
marketplace pricesells slowly62% of the probability falls on outcomes that sell fewer than 21 times a week (3 a day): at the lowest price they won't be bought right away.rare inputsThe rarest input sells 2 times a day. Buying them up will take long, and the price will change meanwhile.
profit on prices without sales at least −117 ₽rare inputsThe rarest input sells 18 times a day. Buying them up will take long, and the price will change meanwhile.
profit on prices without sales at least −380 ₽rare inputsThe rarest input sells 0 times a day. Buying them up will take long, and the price will change meanwhile.