CS.MONEY vs CSFloat

Price aside, the sharpest difference between the two is how much of the item each one describes. CS.MONEY publishes an unusually detailed item model. CSFloat publishes less raw metadata but adds two things that change what you can do with it: its own predicted price and a real permalink to the individual unit.

Those field differences are not cosmetic. Explicit phase naming is the clearest example: Emerald and Ruby do not trade at the same money as the numbered phases, so any price view that lumps all Dopplers together is quoting you an average of items you would never swap for each other. Full-precision float matters if you are chasing an extreme low or high wear, where two listings that both round to the same displayed number are worth very different amounts. And a broken-out sticker or charm value answers the question people most often get wrong on a stickered item: whether you are being asked to pay for the skin, or for the skin plus somebody else's application costs.

Two quotes for the identical weapon, wear and pattern can both be fair. What differs is everything wrapped around the item.

• Seller fees vary enormously across the market. Verified rates run from 2% on CSFloat and 2.5% on BUFF163 up to 12% on Skinport and 15% on the Steam Community Market. A seller pricing to net the same amount will post very different numbers depending on which of those they are standing in. CS.MONEY's own fee structure is not in that verified set, so check it on CS.MONEY's fee page rather than assuming a figure — anyone quoting you one from memory is guessing.
• Stock costs money to hold. A price attached to a published delivery speed is a price for immediacy: someone has already taken the item off the market, carried the risk of it moving, and is selling you the certainty of getting it now. A peer listing on CSFloat has no such carrying cost, but also no guarantee of speed — it sits until it sells.
• Trade locks split the market in two. Steam's post-trade hold means an item you cannot move for days is genuinely worth less than an identical unlocked one, and the gap widens on volatile items. This is exactly the axis a published delivery speed speaks to, and it is a real reason two quotes diverge rather than an accounting quirk.
• Payment rails are not interchangeable. Money in as card, crypto or a topped-up site balance is not the same as money out as a payout, and a "price" you can only spend as store credit on the same site is a discount voucher wearing a price tag. Deposit bonuses, conversion spreads and card charges all land on your side of the trade.
• Regional and account access is a real constraint. Trade holds, mobile authenticator requirements, restricted or banned accounts and payment methods that simply do not work in your country all determine whether a quoted price is available to you specifically.
• Rare attributes are thin markets. For an unusual seed, an extreme float or a Ruby, there may be one or two units in existence for sale anywhere. One listing sets the price, and it can move a long way on a single sale.

The useful question is not "which site is cheaper" but "which of these prices could I realise". A price you cannot convert back into money you want, on a market you are allowed to sell into, from an account you actually have, is not a price. Three tests separate the two:

• The currency you would be paid in. Cash out, site balance, or credit toward another skin — these are three different assets. Compare like for like or you will conclude something is cheap when it is merely illiquid.
• The market you can actually sell into. If your exit route is a market where the seller fee is 12% or 15%, that cost belongs in your entry decision. Buying at a small discount into an expensive exit is a loss dressed as a saving.
• The account you would need. Verification, region, trading eligibility and Steam-side restrictions decide whether the best-looking quote is reachable by you at all.

The practical fix is to pick one reference and convert everything else into it, rather than trusting each site's own suggestion about its own inventory. CSFloat's Final is a defensible yardstick for that job because it is a per-item prediction on a market with real permalinks — you can click through and inspect the exact unit it was computed on, which you cannot do with a blended market average. This is the approach CSWatcher takes: it scores listings on other markets against CSFloat's Final rather than against each market's own number, and on CS.MONEY it is alert-only — it points you at the listing and you buy it yourself. Final is still a prediction, and it is least reliable exactly where the money is strangest: heavily stickered items, extreme floats, one-of-a-kind seeds. Treat it as a yardstick, not a verdict.

Where it wins

If the attribute you care about is one CS.MONEY publishes and others do not, it wins on information alone. Buying a Doppler where the phase is named explicitly — Emerald and Ruby included — removes the single most common way people overpay in that category. The same applies to full-precision float chasing, and to stickered or charmed items where the applied value is broken out instead of buried in the ask. Add a published delivery speed and it is a strong option when you want a specific item now and are willing to pay for the certainty. And for genuinely thin supply, it sometimes simply has the unit and nobody else does, which settles the comparison.

Where it doesn't

If you intend to resell reasonably soon, you need a checkable exit, and that favours a market with per-listing permalinks and a published prediction you can audit against real units.

There is also a mechanical caveat worth knowing as a buyer, because it affects every third-party price view you will ever see quoting CS.MONEY. CS.MONEY throttles automated reading hard and silently. It refuses without a retry hint and without a captcha, and a refusal persists for a long stretch afterwards while the site still looks perfectly healthy in a normal browser tab. The practical consequence: any aggregator, price checker or comparison table showing you CS.MONEY numbers may be working from data that is quietly stale, with no visible warning that it is. Treat CS.MONEY figures on third-party surfaces as indicative and confirm on the site itself before you commit. The absence of a published listing timestamp compounds this — you cannot tell whether something appeared a minute ago or has been sitting unwanted for a week, and listing age is information that usually tells you whether a price is real.

How CSWatcher watches CS.MONEY — what it reads, which rules fire there, and why a quiet stretch on that lane is the design working rather than a fault — is covered at /marketplaces/csmoney; note that every market, CS.MONEY included, scans and alerts on Spectator, the free plan, and that both surfaces cover this market identically. CSWatcher is not affiliated with or endorsed by CS.MONEY, and is not affiliated with or endorsed by CSFloat.