Set a price target, a discount, a rare pattern. Prefire — the CSFloat sniper and auto buy bot — watches the market around the clock and buys the moment a match lists. ★ NEW: standing Buy Orders by paint seed & float with Auto-Update — always #1 in the order book, always at the lowest possible price.
Standing orders that live on CSFloat's servers and can pin exact paint seeds, community tier lists and float brackets — filled in milliseconds, 24/7, even with your PC off. Auto-Update keeps every order #1 at the lowest possible price against the real competition for your exact seeds: it raises you by CSFloat's minimum step when outbid, trims you down when you're overpaying, freezes out bidding wars — and with Confirm mode, proposes every edit for your approval first. Every plan includes it: Free tastes 1 simple order · Starter 3 · Pro adds Auto-Update · Max unlocks seeds & floats.
Two-step Arm → Start, price ceilings, glitch-price refusal and a first-cycle safety stop. Nothing runs, triggers or buys until you say so.
Flag any item under your dollar target, or below its predicted price by a discount you choose — with instant alerts and saved setups.
The hands-off buyer: the instant a match appears, Prefire purchases it in seconds, no click needed. Auto-Bargain (binding max-discount offers) is in Beta — locked until fully tested.
The windfall engine. Hunt rare patterns and float-capped crafts the market lists as ordinary skins — Case Hardened blue gems, fade percentages, Doppler phases, Crimson Webs — the snipes that turn $50 into serious money.
Relentless by design: re-arms itself after every hit and keeps sniping listing after listing, completely hands-off.
Watch entire searches — a colour, a collection, a sticker craft — and the freshest listings the second they appear. Nothing gets past it.
Two ways to run it. We recommend the browser app — nothing to download, any OS — plus the free Bridge extension; or the Windows app if you would rather close your browser and keep sniping. Sign in and your plan loads automatically — no key to copy.
Type weapon · skin · wear, or sync your CSFloat favourites. Set an under-$ target or a deal % per item.
Two-step Arm → Start. The engine scans continuously and the full log is waiting for you in the morning.
Prefire is an automated CS2 (Counter-Strike 2) skin sniper for the CSFloat marketplace. It scans CSFloat 24/7 and auto-buys underpriced skins the moment they appear, based on the price targets, discounts, rare patterns and float ranges you set.
Add the skins or whole searches you want, then set a trigger — a dollar price, a percentage discount below CSFloat's predicted price, a rare pattern or a float range. Prefire scans CSFloat continuously and, when a listing matches, alerts you or buys it hands-free with Auto-Buy.
Auto-Buy has multiple safety layers: a hard spend ceiling (default $10 so you test with small buys first), a typed ARM confirmation, a first-cycle safety stop, and automatic rejection of glitch prices. You can cancel any purchase directly on CSFloat (which charges its own 4% cancellation fee), and you're responsible for the rules you set. The default request cadence is tuned to stay within CSFloat's usual limits to minimize the risk of a softban, though no setting can guarantee it.
Yes. Prefire covers the full Counter-Strike 2 (formerly CS:GO) skin catalog on CSFloat — knives, gloves, rifles, AWPs, pistols, stickers and rare pattern crafts.
There's a free plan with a 24-hour live demo. Every paid plan includes hands-off Auto-Buy: Starter is $19/mo (rules, instant alerts, Auto-Buy), Pro is $79/mo (adds pattern & float filters and continuous runs), and Max is $99/mo — everything, including Filter Sniping across entire CSFloat searches and Newest-listings sniping. New users can claim a free 24-hour Max trial on Discord.
For live sniping, yes. Prefire reads CSFloat from your own connection, so the desktop app — or the browser tab — needs to be open while the engine watches. Close it and sniping pauses; nothing is bought while you are away. The exception is Buy Orders: those are standing orders that live on CSFloat's own servers, so they keep filling 24/7 even with your PC off.
Almost. The browser app needs no download, but it does need one free Chrome or Edge extension — the Prefire CSFloat Bridge. That is not a technicality we can design away: CSFloat serves its data only to its own site, so a web page cannot read it, and requests coming from a shared server get rate-limited. The extension performs each CSFloat request from your own browser and your own connection instead. If you would rather install nothing in your browser, use the Windows desktop app.
It is the piece that lets Prefire read CSFloat as you rather than as a server. It only ever calls csfloat.com — pinned independently in both the extension and the backend — it holds no Prefire password or license key, and it stops working the moment you close the Prefire tab. On the desktop app it does one extra job: reading the competing bids on your standing buy orders from your logged-in CSFloat session.
Yes, through the browser app — it runs in any Chrome or Edge browser on macOS, Linux or Windows, with the Prefire CSFloat Bridge extension installed. The downloadable desktop app is Windows-only today.
Float sniping is buying a CS2 skin for its rare float value — very low (cleaner than normal) or range-capped for a specific look — before other buyers spot it. Because float changes a skin's price, a mispriced low-float listing can be worth far more than its asking price. Prefire watches float ranges on CSFloat for you and alerts or buys the moment one lists.
CSFloat's own docs say automated actions on your account are at your own risk, and rate limits apply per endpoint. Prefire ships with a default request cadence tuned to stay within CSFloat's usual limits to minimize softban risk — no tool can guarantee it. You stay responsible for your account and the rules you arm.
Compare a listing's price against CSFloat's own predicted price rather than shopping by sticker price alone — a $40 skin at 20% under predicted is a better deal than a $10 skin at a fair price. Prefire automates that comparison: set a discount-percentage rule and it flags or buys anything that lists below your threshold, across as many items — or, on Max, whole searches — as your plan allows.
Yes — it's the flagship. Every plan includes the Buy Orders module: standing CSFloat orders with typed confirmations and wallet-tracking price ceilings. On Max you can pin exact paint seeds, community tier lists and float brackets, see the real competing orders for those exact seeds, and let Auto-Update keep each order #1 at the lowest possible price (auto-raise when outbid, trim when overpaying, freeze on bidding wars — with an optional confirm step on every edit). Free tastes 1 simple order, Starter 3, Pro adds unlimited simple orders + Auto-Update.
An auto buy bot watches marketplace listings and completes the purchase for you the moment one matches your rules — no notification-to-checkout race. Prefire's Auto-Buy does this on CSFloat with a hard spend ceiling, a typed ARM confirmation and a first-cycle safety stop, so a misconfigured rule costs little.
Yes. Every paid plan fires instant alerts when a listing matches your rules; Auto-Buy is optional and stays disarmed until you explicitly arm it with a typed confirmation, so an alert-only setup can never spend money.
Yes — pattern hunting is what the seed filters are for. Watch exact paint seeds or seed lists per item (Case Hardened blue gems like the AK-47 661 family, Marble Fade Fire & Ice, Crimson Web triple webs, Slaughter shapes), hunt fade percentages, and put price rules on Doppler phase items like Ruby and Sapphire. Prefire's tier-list guides map every family, and the engine watches the seeds around the clock.
Auto-Buy spends real money. It is protected by multiple safety layers — a hard spend ceiling (default $10 so you can test with small buys first), a typed ARM confirmation, a first-cycle safety abort, and automatic rejection of glitch prices. You can still cancel any purchase directly on CSFloat (which charges its own 4% cancellation fee). You alone are responsible for the rules you set and any purchase they make — Prefire does not refund CSFloat fees or purchases.
Built to reduce softban risk: the default request cadence is tuned to stay within CSFloat's usual limits and to minimize the chance of a softban — though no cadence can guarantee it. Higher plans let you tune the cadence — change it at your own risk.
Prefire in a tab. Nothing to download, nothing to keep updated, and no installer to trust. The engine runs on our servers, but every CSFloat request is carried out through your browser by the free Prefire CSFloat Bridge extension — so the market still sees your connection, not ours.
The whole engine runs on your PC as its own small app. No browser required, no extension required, and it keeps working while you use the rest of your computer. Pick this one if you want to close your browser and leave the engine running.
Two things stand in the way of a pure web app, and neither is ours to change. CSFloat serves its data only to its own site, so a page on another domain is not allowed to read it. And requests arriving from one shared server address get rate-limited no matter whose account they belong to. A browser extension is the one component blocked by neither: it can call csfloat.com directly, and the call leaves from your machine on your connection. That is the whole job it does.
The desktop app has neither problem, which is why it needs nothing extra.
WHAT THE EXTENSION CAN AND CANNOT SEE →Your licence is the account, not the device. Sign in anywhere and the plan follows.
If sniping is already running elsewhere, the second surface pauses itself and offers to take over — so two sessions can never buy the same listing twice.
Standing orders live on CSFloat's servers, so they keep filling around the clock whichever app you use — even with everything closed.
The Chrome Web Store listing is in review. Until it's approved you can load the same extension by hand — it's the identical build, just installed manually instead of in one click.
chrome://extensionsChrome shows a "disable developer mode extensions" reminder while it's loaded this way — that's normal for manual installs and disappears once you switch to the Store version. When the Store version lands, remove this one first so you don't run two copies.
The Chrome Web Store listing is in review. Until it goes live, install it directly — it takes about half a minute and works exactly the same.
Open chrome://extensions and switch on Developer mode, top right.
Click Load unpacked and pick Prefire's extension folder. Desktop users: the app shows you the exact path with a copy button, in the Buy Orders builder.
Reload Prefire. Sign in to csfloat.com in the same browser and you are done.
SIMPLE SNIPING watches one exact item per line — precise, light, and you can add a second rule to the same item. FILTER SNIPING (Max) watches a whole search instead: any Doppler (Ruby) knife under $500 · any Agent at 20% below predicted · anything from the Kilowatt collection under $5 in FN · any weapon carrying a Titan (Holo) | Katowice 2014 sticker. Both run side by side — Start runs everything you've armed; Stop freezes it all.
UNDER $ — a hard dollar target, set slightly under the recent floor to catch mispriced listings the second they appear. DEAL ≥ % — relative to CSFloat's own predicted price; start at 8–12%. Combine both on one rule: either condition triggers. Apply-to-all arms a big watchlist fast; saved setups snapshot items + rules + filters.
Real money — off by default, double-confirmed, audited in Sniper Health. Auto-Buy never fires without a price rule, never pays above the rule's target, and never above the global ceiling in Sniper Settings — set both. Auto-Bargain sends the maximum-discount offer instead of buying; an accepted offer is binding — it is currently in Beta and locked while the offer flow is tested.
PATTERN TIER — curated lists of famous rare patterns (Case Hardened blue gems, high fades, web patterns) caught while they're still priced like ordinary skins. EXACT PATTERN — grail-hunt one specific seed. FLOAT RANGE — float-capped crafts and top-of-wear resale. Watching the newest listings is a Max exclusive.
Pick your surface on the get-the-app page. We recommend the browser app — it runs on any OS with the free CSFloat Bridge extension and there is no installer to download; the Windows desktop app is there if you would rather close your browser and leave the engine running. Sign in with the email and password you used here — your plan loads automatically, no key to copy. Either way every CSFloat request leaves from your own connection, never a shared server — the safest pattern for your account. Keep the app (or the tab) open while you want the engine watching; close it and sniping pauses. It only talks to the market and your account — never your files or anything else on your computer.
Type weapon · skin · wear ("karambit doppler fn") or press Sync CSFloat favourites to import everything you already watch. Each favourite gets its own trigger rule — an under-$ target, a deal %, or both.
Two deliberate steps: Arm locks the board, Start runs the engine. Stop freezes everything instantly. Every hit lands in Triggered Now with a direct Buy link — and the full log is in Sniper Health in the morning.
Release notes are for account holders. Creating an account is free and takes a moment.
By creating an account you agree to Prefire's Terms of Service, Privacy Policy, and Refund & Cancellation Policy, and you confirm you are at least 18. Auto-Buy spends real money and automation may breach CSFloat/Steam rules — you use it at your own risk.
Auto-Buy spends real money. It is protected by multiple safety layers — a hard spend ceiling (default $10 so you can test with small buys first), a typed ARM confirmation, a first-cycle safety abort, and automatic rejection of glitch prices. You can still cancel any purchase directly on CSFloat (which charges its own 4% cancellation fee). You alone are responsible for the rules you set and any purchase they make — Prefire does not refund CSFloat fees or purchases.
Built to reduce softban risk: the default request cadence is tuned to stay within CSFloat's usual limits and to minimize the chance of a softban — though no cadence can guarantee it. Higher plans let you tune the cadence — change it at your own risk.
Every skin your armed Auto-Buy has purchased for you, newest first.
Plain-English guides to sniping the CSFloat marketplace — how it works, what to watch for, how to catch underpriced skins, and how to do it safely. Free to read; the tools that automate it start at $0.
Every pattern family that moves prices, organized by item: which seeds and phases matter, how the tiers work, and how to hunt them. Browse the full tier-list hub →
Sniping on CSFloat means catching an underpriced listing in the seconds after it appears — before other buyers. You watch new listings, know the fair price, and buy instantly when something lists below it. Done manually that means refreshing "Most Recent"; a sniper automates the watch–compare–buy loop.
The honest baseline: open CSFloat, sort by Most Recent, and refresh. When a good skin lists below what it's worth, you add it to your cart and check out — or send a bargain offer. The problem is speed and attention. Genuinely underpriced listings are gone in seconds, and nobody can refresh a page 24 hours a day. That's the entire reason sniping tools exist.
A sniper watches new listings for you and matches them against rules you set: a price target (buy under $X), a discount versus CSFloat's own predicted price, a float range, or a rare pattern seed. When a listing matches, it either alerts you or — if you've armed it — buys hands-free. Lighter tools stop at browser price tags or Discord pings and still need you to click; Prefire runs the whole loop on your own connection — as a Windows app or in your browser — buying the moment a match lists.
Two rule types do most of the work. An under-$ target set slightly below a skin's recent floor catches sudden dips. A deal % rule — say 8–12% below CSFloat's predicted price — catches anything mispriced regardless of the sticker amount. Combining both is the sweet spot: a hard ceiling and a relative discount. Start narrow and loosen as you learn where the deals actually appear.
CSFloat's own docs say automated actions on your account are at your own risk, and rate limits apply per endpoint. Prefire ships with a default request cadence tuned to stay within CSFloat's usual limits to minimize softban risk — but no tool can guarantee it, and you stay responsible for your account and the rules you arm. Auto-Buy also spends real money: it has a hard spend ceiling (default $10 so you test with small buys first), a typed ARM confirmation, and a first-cycle safety stop.
CSFloat has native Buy Orders for some filters, but no general "alert me when anything under my price lists" across price, discount, float and pattern together — which is what a dedicated sniper watches for and, optionally, buys.
Underpriced listings on a busy market can sell in seconds. Manual refreshing rarely wins them; an automated watcher that fires the moment a match appears is the only reliable way to be first.
Not for live sniping — Prefire reads CSFloat from your own connection, so the desktop app or the browser tab has to be open while the engine watches. Close it and sniping pauses. Standing Buy Orders are the exception: they sit on CSFloat's own servers and keep filling with everything closed.
Float sniping is buying a CS2 skin for its rare float value — very low (cleaner than normal) or range-capped for a specific look — before other buyers spot it. Because float changes a skin's price, a mispriced low-float listing can be worth far more than its asking price.
Every CS2 skin carries a float from 0 to 1 that decides how worn it looks. That number maps to a wear tier — the lower the float, the cleaner (and usually pricier) the skin:
| Wear | Float range |
|---|---|
| Factory New | 0.00 – 0.07 |
| Minimal Wear | 0.07 – 0.15 |
| Field-Tested | 0.15 – 0.38 |
| Well-Worn | 0.38 – 0.45 |
| Battle-Scarred | 0.45 – 1.00 |
Two things make float matter. First, the top of a tier: a 0.00–0.01 Factory New looks noticeably cleaner than a 0.06, and collectors pay a premium for it. Second, float caps: some skins can't roll below a certain float, so an unusually low one is genuinely rare. Sellers often price by wear tier alone and miss these — which is exactly the gap a float sniper exploits.
Set a float range plus a price rule on the skins you care about. The engine checks every new listing's float against your range and, when one matches, alerts you or buys it. On Prefire, float and pattern filters are part of the Pro plan — you can hunt a specific low-float window across an item and let the engine watch it around the clock.
It depends on the look you want: the lowest floats in a wear tier (e.g. a 0.00–0.02 Factory New) command a premium, while some players deliberately want a specific range to show more of a pattern. "Good" means rare for that skin and priced below what that rarity is worth.
Less so — on inexpensive skins the float premium is small. Float sniping pays off most on higher-value skins and collections with float caps, where a clean or capped float is worth a real multiple of the ordinary price.
Some skins can't roll below a minimum float by design, so their "Factory New" still shows some wear. A listing under that expected floor is unusually rare — a prime float-sniping target.
The trick isn't a lower sticker price — it's finding skins listed below what they're actually worth. Five ways to do it: compare against the predicted price, watch the newest listings, set discount alerts, send bargain offers, and buy in off-peak windows.
A $40 skin at 20% under CSFloat's predicted price is a better buy than a $10 skin at a fair price. Shop by discount, not by absolute cost — that's how you find value instead of just cheap.
Underpriced skins usually appear as fresh listings from sellers who price by wear alone. Watching "Most Recent" — or letting a tool watch it for you — is where the real deals surface first.
Instead of refreshing, set a rule: alert me when this skin (or any skin in a search) lists more than X% below predicted. Prefire fires the alert — or buys — the moment a match appears.
Many listings accept offers below asking. A standing bargain rule sends a max offer automatically; some get accepted at a real discount. Remember an accepted offer is a binding purchase.
Prices soften at predictable times — mid-week lulls and the dip after a big case or operation release, when supply spikes. It's not a guarantee, but patient buyers catch better prices away from weekend peaks.
Doing all five by hand is a full-time job. Set your rules once and let the engine watch — the free plan already lets you scout the live market, and paid plans add alerts and hands-off Auto-Buy across as many items, or whole searches, as your tier allows.
Generally mid-week and just after a major release, when new supply pushes prices down, rather than at weekend peaks. Timing helps at the margin; catching a specific underpriced listing matters more, which is where alerts and sniping come in.
CSFloat is an established CS2 marketplace with its own escrow and predicted pricing. As with any market, use your own account responsibly — and note Auto-Buy spends real money, so start with a low spend ceiling.
Listed below the fair market value for that exact skin — its wear, float and pattern — usually measured against CSFloat's predicted price. A cheap sticker price isn't the same as a good deal.
Every CS2 skin rolls a paint seed from 0 to 999 that decides how its finish sits on the weapon. A handful of seeds — Case Hardened blue gems, full fades, centered crimson webs — are worth multiples of the ordinary price. Pattern sniping is catching those seeds listed at ordinary prices.
Most sellers list by wear and price, not by seed. Unless someone has checked the pattern against a known tier list, a rare seed can sit at the normal price until a collector notices — and then the market reprices it upward. That lag is the whole opportunity.
A few examples of what pattern hunters chase (facts, not price promises):
There are two ways to hunt. A pattern tier targets a curated list of good seeds for a skin — broader, more matches. An exact seed hunt watches for one specific grail pattern — rarer, but the biggest payoff. Prefire's scan filters (Pro) support both: pick a tier or an exact seed, add a price ceiling, and let the engine watch for it.
A number from 0 to 999 assigned when a skin is created. It determines the exact placement of the pattern — where the blue sits on a Case Hardened, how a fade is arranged — and is fixed for the life of that skin.
No. Case Hardened exists on several weapons (the Five-SeveN, Karambit and more), each with its own sought-after seeds. The AK-47's blue gems are simply the most famous and valuable.
Yes — set an exact paint seed plus a price rule and the engine watches every new listing for that specific pattern, alerting or buying when it appears at your price.
Honest answer: automating actions on your CSFloat account carries real risk, and no tool can promise zero. CSFloat's own docs say automation is at your own risk and rate limits apply per endpoint. The sensible approach is a conservative request cadence, small spend limits, and staying in control — which is how Prefire ships by default.
A softban is a temporary, automatic throttle a marketplace applies when it sees behaviour that looks automated — too many requests too fast, or patterns that don't look human. It's usually short-lived (minutes to hours) and clears on its own, but while active your requests are rate-limited or rejected. It is not a permanent ban of your account for normal use.
Two things drive it: request volume (hammering the API faster than a person would) and account automation generally, which CSFloat's documentation explicitly says you do at your own risk. Any tool that scans listings or buys for you is making requests on your behalf — so the question isn't "is a tool safe," it's "does it stay within sensible limits, and do you keep control."
Prefire ships with a default request cadence tuned to stay within CSFloat's usual limits, so out of the box it behaves conservatively. Higher plans let you tune that cadence — faster is more aggressive, and that's your call and your risk. On the money side there are layered guards: a hard spend ceiling (default $10 so you test small first), a typed ARM confirmation before anything can buy, a first-cycle safety stop that aborts if rules trigger instantly on stale data, and automatic rejection of glitch prices. None of this guarantees you'll never see a softban — it lowers the odds and keeps a bad rule from spending badly.
You stay responsible for your account and the rules you arm. Start with the defaults, keep the spend ceiling low until you trust your setup, and don't crank the cadence just to shave seconds. If you're not comfortable being responsible for what your rules buy, keep Auto-Buy off and use Prefire for alerts only.
Normal, conservative use is far more likely to hit a temporary softban (a short throttle) than a permanent ban, and the risk rises with aggressive request rates. Prefire's default cadence is tuned to stay within CSFloat's usual limits, but no tool can guarantee it, and you use automation at your own risk per CSFloat's docs.
Softbans are typically short — minutes to a few hours — and clear automatically. During one your requests are rate-limited, so an armed sniper simply won't fire until it lifts. Backing off the request cadence reduces how often you'd hit one.
Keep the default cadence, set a low spend ceiling, arm one well-tested rule at a time, and watch the first runs before trusting it. Prefire's ARM confirmation, first-cycle safety stop and glitch-price rejection are there so a mistake costs little.
"CSFloat sniper" covers three very different kinds of tool: Discord bots that ping you, web scanners that list deals on a dashboard, and desktop snipers that complete the purchase. Full disclosure — Prefire is our product. The comparison below sticks to what each category actually does, as of July 2026.
Discord alert bots (e.g. CS2 Market Sniper) watch prices and post a ping in a channel. They're free or cheap — but every subscriber gets the same ping at the same moment, and you still have to open CSFloat, find the listing and check out before the rest of the channel does.
Web deal scanners (e.g. CS2Locker) compare live listings against reference prices and show hits on a dashboard. Free tiers are typically capped and delayed — CS2Locker's own pricing shows free users a limited feed about a minute behind paying ones — because with deals, speed is the entire product. You still complete every purchase yourself.
Auto-buy snipers execute the deal: when a listing matches your price, discount, float or pattern rule, the engine buys it hands-free, with spend guards deciding what it's ever allowed to spend. Prefire is in this category; cs2sniper.org runs an invite-only closed beta in the same space.
| Discord alert bot | Web deal scanner | Prefire | |
|---|---|---|---|
| On a match | Ping in a channel | Row on a dashboard | Buys it hands-free — or alerts, if unarmed |
| Who gets the deal | Every subscriber at once | All subscribers; free tiers delayed & capped | You — your rules run privately |
| Where it runs | Shared cloud | Vendor's servers | Your own PC, your own connection |
| Spend safety | — (you buy manually) | — (you buy manually) | Spend ceiling, typed ARM, first-run stop, glitch-price rejection |
| Cost | Free / cheap | Free tier + paid speed | Free plan; every paid plan includes Auto-Buy, from $19/mo |
Competitor details are from their public pages as of July 2026 and can change — check the current pages before you pay for anything, including ours.
A genuinely underpriced listing on a busy market is gone in seconds, and an alert leaves the slowest step to you: noticing the ping, opening the page, and checking out faster than everyone else who got the same ping. The race is decided in the second after the listing appears, not the minute after the notification. That's why serious flippers end up on execution tools. If you want to see the difference before paying anything, the free plan lets you scout the live market, and our Discord posts a delayed feed of what the engine actually caught.
Yes, within limits: a watcher that checks new listings around the clock will catch mispriced skins a human refreshing the page won't. The differences that matter are whether it buys or only alerts, how fast it reacts, and whether its request pattern stays within CSFloat's limits.
A scanner shows you deals it found — you still have to be there, click, and check out before anyone else. A sniper executes: when a listing matches your rule it completes the purchase hands-free. On a fast market that difference is usually the whole deal.
CSFloat's docs say automated actions on your account are at your own risk, with per-endpoint rate limits. Whatever tool you pick, a conservative request cadence and small spend limits are what keep the risk manageable — no vendor can honestly promise zero.
Flipping is buying a skin below its fair value and reselling it at that value. No prediction, no gambling — if you bought right, the profit was locked in the moment you bought. The whole game is knowing what "right" means and getting there before someone else does.
Say a skin's fair value — CSFloat's predicted price for that exact wear and float — is $100, and a seller lists at $85. You buy, relist at $97 (slightly under value, so it sells this week instead of this month), and pay the marketplace's seller fee on the way out — low single digits on CSFloat (tier-dependent; check your account's current rate). Net: roughly $10 on $85 committed, about 12% per flip. The same flip sold on the Steam market loses ~15% in fees and traps the proceeds as wallet credit you can never withdraw — which is why flippers sell where fees are low and money stays money.
A 30% margin on a skin nobody searches for is a donation to your own inventory. Before buying, check how often that exact skin actually sells: popular rifles, AWPs and mid-tier knives in the $20–$300 band turn over in days, while niche grails can sit for weeks no matter how good your price was. The working rule: a smaller margin on something that sells this week beats a bigger margin on something that might sell this quarter.
Underpriced listings exist because sellers price by wear tier alone, need cash fast, or never checked the pattern. Four places they surface: the newest listings feed, where mispricings are visible for seconds; discount-vs-predicted gaps (a listing 10%+ under the predicted price is mispriced by definition); float and pattern value the seller didn't price in — see the float sniping and pattern sniping guides; and bargain offers on listings that accept them. Catching any of these by hand means being there first, every time — which is the part worth automating. Set the rule once and let the engine watch.
Flip with money you can leave parked: from buy to sold-and-cleared, your capital is committed — and hold windows before an item can be resold make that commitment longer than beginners expect. Keep a hard per-buy ceiling so one fat-fingered rule can't drain the bankroll, and track every flip — buy price, sell price, fees, days held — because the numbers, not the highlights, tell you whether you're actually profitable. Prefire's Auto-Buy enforces the ceiling and logs every purchase to your history automatically.
Prices move while you hold — a patch, a case release or a market swing can compress your margin to zero. Automation on your CSFloat account is at your own risk per their docs (the softban guide covers this honestly). And it's real money: start small, prove your read of the market on cheap flips, and scale what works.
Less than most people think — margins are percentage-based, so a $50–100 bankroll teaches you the full loop: find, buy, relist, wait out any hold window, sell. Scale up only after several flips have gone the way you predicted.
Yes, but the easy edges are gone: obvious underpricing gets bought within seconds, usually by automation. Profit now comes from selection (knowing what's actually mispriced) and speed (being first to it) — not from browsing longer than the next person.
On liquid mid-tier skins, single-digit percentages after fees are normal; the bigger wins come from mispriced floats, patterns and panic listings. Anyone promising a fixed daily profit is selling something — treat every flip as its own trade.
There are two ways to be first to a deal on CSFloat: a standing buy order that fills automatically at your price, and a watcher that hunts value the moment it lists. They solve different problems — and the strongest setups run both.
A buy order is a standing bid: you name the item and the most you'll pay, and when a matching listing appears at or under that price, it fills — instantly, server-side, with your PC off. It's the cleanest way to buy a skin you already know you want at a price you already know is fair. The catch is competition: orders compete on price, so a better-priced standing order gets the fill before yours.
Fair value moves; your order doesn't. An order is a number you set last week — the predicted price is live. When a $100-predicted skin lists at $80, that's a 20% mispricing, and it sails straight past every order parked at $70. A standing order is also visible pressure on the market: sellers can price just above the best order, which means the deals that fill your order are, by construction, rarely the mispriced ones. And the biggest mispricings — a clean float or a grail pattern the seller never checked — are mispriced relative to what the skin is actually worth, which no fixed number expresses.
| You want | Use |
|---|---|
| This skin, whenever it's under $Y | Buy order |
| Anything 10%+ under its predicted price | Sniper rule (deal ≥ %) |
| A grail seed or float window at an ordinary price | Sniper (pattern / float filters) |
| A guaranteed fill at a known price, PC off | Buy order |
| A whole CSFloat search watched around the clock | Sniper (Filter Sniping) |
Prefire's Buy Orders module closes most of the classic gaps. On the Max plan an order can pin exact paint seeds (or a whole community tier list — blue gems, Fades — auto-split at 25 seeds per order) and exact float brackets, so "a grail pattern at my price" IS expressible as a standing order that fills in milliseconds. The module shows the real competing orders for those exact seeds — not the generic book — and Auto-Update (Pro+) keeps every order first at the lowest possible price: it raises you by CSFloat's own minimum increment when outbid, trims you back down to just above #2 when you're overpaying, freezes the order instead of chasing a bidding war, and — with Confirm mode on — proposes every edit for your one-click approval instead of acting alone. Wallet-tracking ceilings (per-order = your CSFloat balance, total = 10×) bound everything it does.
Orders are your floor: commodity skins you'd always take at a known price — and, with seed & float orders, your grails too. The sniper is your hunter: it evaluates every new listing against rules a fixed bid can't express — discount against the live predicted price, best-deal judgment calls, entire searches — and either alerts you or buys within the guards you set (a hard spend ceiling, a typed ARM confirmation, and a purchase history you can audit). The flippers who catch the most don't pick a side; they let orders handle the describable and the sniper handle the judgment calls.
Only at or below the fixed price you set on that item. If a skin is worth $100 and lists at $80, a $70 buy order sleeps through it — a sniper watching for "15% under predicted" catches it. Orders guarantee a price; they don't hunt for value.
Usually one of three things: a higher standing order had priority, the listing was bought outright before your order matched, or your criteria were narrower than you thought. A standing order also shows your hand — sellers can price just above it.
Both. An order is perfect for a commodity skin you'd always buy at a known price. A sniper covers what a fixed price can't express — a discount relative to the live predicted price, float windows, pattern seeds, whole-search hunting — with your own spend guards on every buy.
An auto buy bot watches marketplace listings and completes the purchase for you the instant one matches your rules. No alert, no reaction time, no checkout race — the whole point is removing the human from the seconds that decide who gets the deal.
Three jobs, in a loop: watch every new CSFloat listing for the items or searches you care about, match each one against your rules — a price target, a discount versus the live predicted price, a float window, a pattern seed — and buy the moment something qualifies. Alert-only tools stop at step two and page you; an auto buy bot finishes the job while the notification would still be in transit. On a contested deal, that difference is the deal.
The three automation levels solve different problems. Alerts keep you in control: the bot spots the deal, you decide. Buy orders are standing bids at a fixed price — great for commodity skins, blind to mispricing. Auto-Buy is the full loop: it evaluates value the way you would (discount, float, pattern) and executes the way an order does (instantly, hands-free). Prefire runs all three, so the honest answer to "which one?" is usually a mix.
Hands-free spending needs hard limits, not good intentions. Prefire's Auto-Buy ships with a hard spend ceiling (default $10, so your first runs are cheap experiments), a typed ARM confirmation — the engine won't spend a cent until you deliberately arm it — a first-cycle safety stop so a freshly-armed rule can't go on a spree, and glitch-price rejection that refuses listings priced absurdly below reality (usually stale or bait data). Everything it buys lands in an auditable purchase history.
Start narrow: one skin you know well, a price you'd take without thinking, and the default ceiling. Watch a few cycles in alert mode first, then arm. When the bot's picks match what you'd have bought yourself, widen the rule — more items, a discount trigger, a float window. Automation should earn trust the same way a human buyer would: with small, checkable decisions first. And per CSFloat's own docs, automation on your account is at your own risk — a conservative cadence (Prefire's default) is what keeps that risk low.
CSFloat's docs say automated actions on your account are at your own risk, with per-endpoint rate limits — the same terms that apply to any automation. A conservative request cadence and small spend ceilings are what keep the risk manageable; no tool can honestly promise zero.
You can cancel the purchase directly on CSFloat (which charges its own 4% cancellation fee), and the damage is bounded before that: the spend ceiling caps what any rule can commit, and the first-cycle safety stop means a brand-new rule can't go on a spree the moment you arm it.
The engine checks new listings continuously and sends the buy the moment one matches an armed rule — no notification-to-checkout race, no human reaction time. On contested deals that gap is usually the whole difference between winning and watching.
A listing alert tells you the second a CS2 skin matching your criteria appears on CSFloat — price, discount against predicted, float range or pattern seed. It turns "refresh and hope" into "get pinged and decide."
Underpriced listings on a busy market are gone in seconds, and the person refreshing "Most Recent" is competing against watchers that never blink. An alert engine flips the odds: it does the refreshing around the clock and only involves you when there's a decision worth making. You stop paying attention-tax on the 99.9% of listings that don't matter.
A useful alert is a rule, not a keyword. The four that do the work: an under-$ price target on a specific skin; a deal % — anything listing more than, say, 10% below CSFloat's own predicted price; a float window for clean or capped floats; and a pattern seed (or seed list) for blue gems, fades and webs. Prefire evaluates every new listing against all your rules at once and fires the moment one matches.
Speed after the ping matters as much as the ping. Keep the CSFloat tab signed in, know your cancel rights (CSFloat charges a 4% cancellation fee), and decide your ceiling before the alert — the deals worth alerting on don't leave time for research. If you keep losing races between the alert and your click, that's the signal you've outgrown alert-only.
Alert-only is the right way to start: it costs nothing extra, spends nothing, and teaches you which of your rules actually find value. When a rule has proven itself — the alerts it fires are deals you'd have taken every time — arming Auto-Buy on that rule just removes your reaction time from the equation, inside a hard spend ceiling and with a typed ARM confirmation. Prefire runs both modes side by side, per rule.
An alert fires within the engine's scan cycle of the listing appearing — seconds, not minutes. That's usually fast enough to win uncontested deals manually; for contested ones, hands-off Auto-Buy removes the reaction-time race entirely.
No — instant alerts are part of every paid plan from Starter up, and so is Auto-Buy (having the engine complete the purchase for you). Higher tiers add pattern & float filters (Pro) and Filter Sniping across whole searches (Max).
Yes. Auto-Buy stays disarmed until you explicitly arm it with a typed confirmation, so an alert-only setup can never spend money. Many users run alert-only first and arm Auto-Buy once they trust their rules.
A "blue gem" is a Case Hardened skin whose paint seed rolled a playside that's almost entirely blue. A few hundred of the 1,000 possible seeds show meaningful blue; a handful are grails worth many multiples of an ordinary Case Hardened — and they all look identical in a listing until someone checks the seed.
Case Hardened's finish is a random wash of blue, purple and gold, and the paint seed (0–999) fixes exactly where each color lands. Collectors grade by blue percentage on the playside — the side you actually see in-hand. A seed with a near-solid blue playside is a blue gem; mostly-gold rolls have their own following as gold gems. Wear matters too: the same seed in Factory New shows its colors far better than in Battle-Scarred.
The community has mapped every seed, and a few names come up constantly. On the AK-47, pattern 661 is widely regarded as the #1 playside blue gem, with seeds like 955 and 670 close behind in most tier lists. On knives, the Karambit pattern 387 is the most famous blue gem in the game — its owner has reportedly declined seven-figure offers. Case Hardened also rolls on the Five-SeveN, Bayonet, Flip Knife and more, each with its own sought-after seeds and its own collectors.
Community tier lists rank seeds from Tier 1 (near-full blue playside) down through partial gems, and they're the reference collectors actually trade against. Read them carefully: a seed can be a playside gem and a backside dud, or the reverse, and prices follow the playside. When you evaluate a listing, open the inspect link and look at the side that shows — the seed number alone doesn't tell you which face got the blue.
Every CSFloat listing exposes its paint seed, which makes blue gems a search problem: the skin is valuable because of a number most sellers never check. That's the whole sniping opportunity. With Prefire you can watch an item for an exact seed or a list of seeds — alert-only or Auto-Buy — and browse the whole Case Hardened family as one group instead of weapon by weapon. When a tiered seed lists at an ordinary Case Hardened price, you want to be the one who noticed first.
Pattern 661 is widely regarded as the #1 playside blue gem — a near-solid blue magazine-side top. Seeds like 955 and 670 sit just behind it in most community tier lists. "Best" is playside-driven: the blue you see in-hand is what collectors pay for.
They're a collector market: prices are driven by scarcity and demand for specific seeds, and liquidity is thin — a top-tier gem can take a long time to sell at its ask. Treat any blue gem buy as a collectible position, not a guaranteed flip.
Every CSFloat listing shows the paint seed and the inspect link, and community tier lists map seeds to blue coverage. A seed filter automates the hunt: watch an item for exact seeds (or a seed list) and get alerted — or auto-buy — when one lists at your price.
Six pattern families decide most of the pattern money in CS2. Each family below gets its own tier-list guide, organized item by item — what the tiers mean, which seeds and phases matter on which weapon, and how to catch tiered patterns listed at ordinary prices.
Two families are seed-driven — Case Hardened and the fades read their look from the paint seed, so the same item name hides wildly different values. Doppler is phase-driven (a different finish index per phase), and Crimson Web, Marble Fade and Slaughter sit in between: seed-driven, but ranked by recognizable features rather than raw coverage. In every family the money logic is identical — most sellers list by item name and wear, a minority of patterns carry the premium, and whoever checks the pattern first wins. Prefire watches seeds and prices for you across all of them: set a seed list or a price rule on the items you collect and get alerted — or auto-buy — the moment a tiered pattern lists at an ordinary price.
Case Hardened rolls a random wash of blue, purple and gold from its paint seed, and every weapon wraps it differently — so each model has its own tier list. This page organizes the family by item: what Tier 1 means, and which seeds the community ranks on each weapon. New to blue gems? Start with what makes a seed a blue gem.
Community tier lists grade seeds by blue coverage on the playside — the side you see holding the weapon. Tier 1 is near-solid blue; each tier down shows more gold breaking through. Backside coverage is tracked separately and trades at a discount. Because tiers are per-model, always check the list for the exact weapon: a grail seed on the AK-47 means nothing on a Bayonet.
The most collected Case Hardened item in the game. Pattern 661 — the famous "scar pattern" — is widely regarded as the #1 playside blue gem, with 955 and 670 close behind on most lists. Top-tier AK gems trade at large multiples of an ordinary Case Hardened, and StatTrak examples multiply again on scarcity.
Home of the most famous pattern skin in CS2: the Karambit 387 blue gem, whose owner has reportedly declined seven-figure offers. Below the grails, the Karambit's curved blade shows blue dramatically, which keeps demand strong across its upper tiers.
The budget entry into blue gem collecting — the slide shows a large, flat playside, so top seeds look striking, and prices sit far below AK or knife gems. A common path: learn the market on Five-SeveN gems, then move up.
Case Hardened rolls on most knife models — Bayonet, M9, Flip, Gut, Butterfly and the newer lines — each with its own seed list and its own collectors. Blade shape decides everything: a seed that pools blue on one model scatters it on another, which is exactly why per-item lists exist.
Every CSFloat listing exposes its paint seed, and most sellers never check it. With Prefire you can watch an exact seed list per item — or browse the whole Case Hardened family as one group — and get alerted or auto-buy when a tiered seed lists at an ordinary price.
No. The finish wraps differently around each model, so a seed that's a top-tier blue gem on the AK-47 can be unremarkable on a knife. Every weapon has its own community tier list — always check the list for the exact model you're buying.
Playside blue coverage. Tier 1 means the visible side is close to solid blue; Tier 2 still shows strong blue with more gold or purple breaking it up. The gap in price between adjacent tiers is usually far larger than the gap in looks.
They trade at a real discount to playside gems — the side you see in-hand drives the premium. A backside gem can still be a smart buy at the right price, but never pay playside-tier money for one.
Fade is the cleanest pattern family in the game: one number — the fade percentage — summarizes how much of the item is covered by the pink-purple gradient instead of yellow. The percentage comes from the paint seed, and each item has its own range and its own price curve.
Most fade items roll from roughly 80% up to a 100% "full fade". The market is steeply top-loaded: the difference between 90% and 95% is modest, while 98–100% is where collectors compete and premiums jump. Percentages are per-item — the same seed produces different coverage on different models — so always read the number for the exact item.
The Karambit Fade is the family's flagship — a full fade is one of the most liquid high-end knives there is. The M9 Bayonet and Butterfly fades follow the same logic with their own ranges, and the newer knife lines (Talon among them) added fresh full-fade chases with smaller supply.
The affordable entry: the Glock-18 Fade is a pistol-priced classic where the top percentages still carry a clear premium. It's the standard first fade — the market is deep and the tier logic is identical to the knives.
The gradient family keeps growing — pistols, SMGs and knife models added over the years each carry their own seed-to-percentage map. The constant across all of them: sellers who list by wear alone, leaving the percentage unpriced. That's the gap a fade hunter lives in.
Set a rule on the fade items you care about and let the engine watch the seeds. When a listing's seed maps to a top percentage and the price says "ordinary fade", you get the alert — or the buy — before a manual browser ever opens the inspect link.
Most fade items roll from roughly 80% up to a 100% "full fade" — the higher the percentage, the more of the finish is covered by the pink-purple gradient instead of yellow. Premiums concentrate in the last few percent: 98–100% is where collectors compete.
The percentage is derived from the paint seed. CSFloat shows every listing's seed and inspect link, and community fade calculators map seed to percentage per item — check the exact model, because the same seed fades differently across items.
Usually, but not automatically — condition still matters, and on some items collectors also chase specific looks (like even color balance) over the raw percentage. A clean high-float full fade can lose to a pristine 95%.
Doppler is the family where the "pattern" is actually the finish itself: each phase is a different paint index, not a paint seed. That one technical detail explains both the huge price gaps between phases and why phase items get mispriced more often than any other family.
At the top sit the solid-color rarities: Sapphire (blue) and Ruby (red), with Black Pearl behind them. Then come the numbered phases 1–4 — visually similar galaxy swirls where Phase 2 usually leads on its pink playside and Phase 4 has its own following for the blue. On Gamma Dopplers the chase finish is the Emerald, above Gamma phases 1–4. Exact ordering shifts per knife; sold listings for your model are the truth.
The classic Doppler knives — Karambit, M9 Bayonet, Bayonet, Flip and Gut — were joined over the years by Butterfly, Talon and most of the newer knife lines, plus the Glock-18 Gamma Doppler on the pistol side. Every model prices its phases differently: a Karambit Sapphire is a grail; the same phase on a smaller knife is merely expensive.
Because the phase is a separate finish hiding under one shared item name. A seller lists "★ Karambit | Doppler" at a phase-blind price, and the gap between phases does the rest. It's the single most common expensive mistake on knife markets — and the easiest one to catch with a watcher that knows what the listing actually is.
Put price rules on the phase items you want and let the engine compare each new listing against what that phase actually trades at. When someone lists a Phase 2 at Phase 3 money — or better — you're first in line instead of hearing about it later.
Sapphire and Ruby sit at the top, followed by Black Pearl, then the numbered phases — Phase 2 usually leads those for its pink playside. On Gamma Dopplers, Emerald is the chase finish. Exact order shifts per knife, so check sold prices for your model.
No — the phase is a different finish (paint index), not a seed. The paint seed only nudges how the finish sits on the blade. That's why phases are listed as distinct items, while blue gems and fades are seed-driven variations of one item.
Because sellers sometimes list by the item name without pricing the phase — and phase price gaps are large. A watcher that flags a cheap listing on a phase item catches exactly this kind of mistake before manual browsers do.
Crimson Web is ranked by something anyone can see at a glance: how many full webs land on the playside, and how centered they sit. Add the family's brutal float behavior — the finish darkens fast with wear — and you get one of the most collectible knife lines in the game.
Two variables: web count and placement. A single full, centered playside web already beats a webless pattern; two clean webs is a serious knife; a triple web is the family's chase pattern. Partial webs and edge placements rank below their centered equivalents on every list.
The red base darkens and the webs lose contrast as float climbs, so condition multiplies the pattern instead of just accompanying it. Factory New examples on the older knives are genuinely scarce — a FN with good webs is a grail, while the identical pattern in Field-Tested is just a nice knife.
The classic Crimson Web knives — Karambit, M9 Bayonet, Bayonet, Flip and Gut — each form webs differently across their blades, so counts and placements that are common on one model are rare on another. The newer knife lines that received the finish extend the family with smaller supply and younger tier lists. (The webbed glove finishes, like Crimson Weave, are their own items with their own patterns.)
Web patterns are visible in the inspect view but invisible in the price column — which is the whole opportunity. Watch the family with seed filters and price rules, and let the engine flag the listing where a multi-web playside is priced like a webless one.
A pattern where three full webs land on the playside — the classic chase pattern for Crimson Web. Web count and how centered the webs sit are what tier lists rank; a single centered web already carries a premium over webless patterns.
Yes — the finish darkens fast with float, and webs lose contrast. Factory New Crimson Webs are genuinely scarce on the older knives, which is why a FN with good webs is a grail while a Field-Tested with the same pattern is just a nice knife.
The glove finishes with webbing (like Crimson Weave) are separate finishes with their own patterns — the classic Crimson Web tier lists cover the knife line. Same idea, different item family and different lists.
Marble Fade swirls red, yellow and blue from the paint seed — and the tier list is about what's missing: the less yellow, the higher the tier. At the top sit the Fire & Ice patterns, where the blade splits into almost pure red and blue.
A Fire & Ice pattern shows a near-clean two-color split with the yellow band minimal or gone. Community lists rank the best examples per knife as "1st max", "2nd max" and so on, by how clean the split is on the playside. On the Karambit they're among the most recognizable pattern skins in CS2, with prices to match.
Below Fire & Ice sit the tricolors — patterns with a strong, deliberate band of all three colors — and then the ordinary rolls where yellow dominates. Tricolors are the budget entry into the family: visibly special, meaningfully cheaper than a max pattern.
The classic Marble Fade four — Karambit, Bayonet, Flip and Gut — carry the famous lists, with the Karambit Fire & Ice as the family flagship. The newer knife models that received the finish have their own max-pattern lists with thinner supply, which is where quiet deals happen.
Marble Fade sellers price by item and wear; the seed decides whether that price is right. Watch the family with seed lists for the max patterns per knife, and take the alert — or the auto-buy — when a Fire & Ice lists at tricolor money.
A pattern where the marble shows almost only red and blue — the yellow band is minimal or gone. Maximum examples ("1st max") are ranked by how clean the two-color split is; on the Karambit they're among the most recognizable pattern skins in the game.
A pattern with a strong, visible band of all three colors — red, yellow and blue — usually ranked below Fire & Ice but above ordinary rolls. It's the budget entry into Marble Fade pattern collecting.
The classic Marble Fade knives — Karambit, Bayonet, Flip and Gut — plus the newer models that later received the finish. The Karambit Fire & Ice is the famous one, and per-knife tier lists rank the max patterns for each model.
Slaughter is the shape-hunter's family: the red swirls occasionally form recognizable figures — hearts, angels, diamonds — and a centered, unmistakable shape on the playside turns a standard knife into a pattern collectible.
The heart is the family's icon, with angels and diamonds as the other established chases. Tier logic is placement-first: a full shape centered on the playside is the top of the list, a partial or edge-of-blade shape ranks tiers below, and everything else is just a clean red knife.
The classic Slaughter knives — Karambit, Bayonet, M9, Flip and Gut — each form their shapes in different spots, so every model has its own pattern list and its own definition of "centered". The newer lines that received the finish added fresh, thinner lists where shapes still go unnoticed.
Premiums are generally smaller than blue gems or Fire & Ice — this is a recognizable-shape market, not a coverage market. The flip side: far more Slaughter listings go out unchecked, so the hit rate on mispriced patterns is better than the headline premiums suggest.
Shape patterns are seed-driven, so the hunt automates the same way as blue gems: keep the community's shape-seed lists on watch per knife, and let the engine alert — or buy — when a heart lists at plain-Slaughter money.
A pattern where the red swirls form a visible heart shape on the playside — the best-known Slaughter pattern, alongside angels and diamonds. Placement decides the tier: a centered, unmistakable shape outprices a partial one near the edge.
The premiums are real but generally smaller — Slaughter patterns are a recognizable-shape market rather than a coverage-percentage market. That also means more listings go out unchecked, which makes mispriced patterns easier to catch.
The classic knife line — Karambit, Bayonet, M9, Flip and Gut among them — plus later models that received the finish. Each model forms its shapes differently, so the pattern lists are per-knife.
These Terms of Service ("Terms") are a binding agreement between you ("you", "your", the "user") and Prefire, the operator of the Prefire software and the https://prefire.trade website (the "Operator", "Prefire", "we", "us", "our"). By creating an account, installing or using the Prefire desktop or web application, or clicking to accept these Terms, you agree to be bound by them and by our Privacy Policy and Refund & Cancellation Policy, which are incorporated by reference. If you do not agree, do not use Prefire.
You must be at least 18 years old (or the age of majority in your place of residence, if higher) and legally able to enter into this agreement. You must also meet the age and eligibility requirements of Steam and CSFloat. You represent that you are not barred from using the service under the laws of any applicable jurisdiction and that you are not on any sanctions or denied-parties list.
Prefire is a tool that you operate. It monitors publicly listed items on the CSFloat marketplace against criteria you define (price targets, discounts, patterns, float ranges, filters) and, when you have explicitly armed the Auto-Buy or Auto-Bargain features, attempts to purchase or make an offer on a matching listing through your own CSFloat account and credentials.
You are responsible for keeping your login credentials, license key, and any CSFloat session information you provide confidential, and for all activity under your account. Notify us promptly of any unauthorised use. You may hold one account unless we agree otherwise, and you may not share, sell, sublicense, or transfer your account, license key, or subscription. We may refuse, suspend, or terminate accounts at our discretion, including for suspected fraud, abuse, or breach of these Terms.
Subject to these Terms and your active subscription, we grant you a limited, personal, non-exclusive, non-transferable, revocable licence to install and use Prefire for your own lawful use. You may not, and may not permit anyone else to: (a) copy, resell, rent, lease, sublicense, distribute, or provide the software as a service to third parties; (b) reverse-engineer, decompile, or attempt to derive source code, except to the extent this restriction is prohibited by law; (c) remove or alter proprietary notices; (d) use the software to build a competing product; or (e) circumvent licensing, request limits, plan gating, or security controls. All intellectual property in Prefire remains ours or our licensors'.
This is important. CSFloat and Steam are independent third parties with their own terms of service, and those terms may restrict or prohibit automated access, scripting, or bot activity. You are solely responsible for reviewing and complying with the terms of CSFloat, Steam, and any other platform you connect to Prefire.
Auto-Buy and Auto-Bargain spend real money. By arming them you acknowledge and agree that:
Paid plans are sold as recurring subscriptions and one-time credit packs, billed through our payment processor, Stripe. By subscribing you authorise us and Stripe to charge your payment method the then-current fees, plus applicable taxes, on a recurring basis until you cancel.
You agree not to use Prefire to: break any law or third-party right; commit fraud, money laundering, or market manipulation; access an account that is not yours; interfere with, overload, or disrupt CSFloat, Steam, or our systems; probe, scan, or bypass security or usage limits; or resell or provide access to others. We may suspend or terminate access for any violation.
Prefire is provided "as is" and "as available", without warranties of any kind, whether express, implied, or statutory, to the maximum extent permitted by law. We specifically disclaim the implied warranties of merchantability, fitness for a particular purpose, title, and non-infringement. We do not warrant that: the service will be uninterrupted, timely, secure, or error-free; it will detect, win, or successfully buy any particular listing; price or float data is accurate or complete; or that use of the service will avoid a softban, ban, loss, or any action by CSFloat or Steam. No advice or information you obtain from us or the service creates any warranty not expressly stated here. Any statement in our marketing, website, or documentation that is not in these Terms is not a warranty, and in the event of a conflict these Terms and the Refund & Cancellation Policy control.
To the maximum extent permitted by law, the Operator will not be liable for any indirect, incidental, special, consequential, exemplary, or punitive damages, or for any loss of profits, revenue, data, goodwill, skins, virtual items, or account access, or for market losses, platform fees, softbans, suspensions, or bans, arising out of or relating to Prefire, whether based in contract, tort, or otherwise, even if advised of the possibility of such damages. Our total aggregate liability for all claims relating to the service will not exceed the greater of (a) the total amount you paid us for the service in the three (3) months before the event giving rise to the claim, or (b) US $100. Some jurisdictions do not allow certain limitations, so some of the above may not apply to you; nothing here limits liability that cannot be limited by law (such as for fraud or death or personal injury caused by negligence).
You agree to defend, indemnify, and hold harmless the Operator from and against any claims, damages, losses, liabilities, and costs (including reasonable legal fees) arising out of or related to your use of Prefire, your purchases and trading activity, your violation of these Terms, or your violation of any law or of the terms of CSFloat, Steam, or any third party.
You may stop using Prefire and cancel your subscription at any time. We may suspend or terminate your access, with or without notice, for breach of these Terms, suspected fraud or abuse, legal or platform-compliance reasons, or discontinuation of the service. On termination, your licence ends and you must stop using the software; sections that by their nature should survive (including Sections 4, 5, 6, 9, 10, 11, 13, and 14) survive. We may modify these Terms; if a change is material we will take reasonable steps to notify you (for example, by email or an in-app or website notice), and your continued use after the change takes effect is acceptance.
These Terms are governed by the laws of the jurisdiction in which the Operator is resident, without regard to its conflict-of-laws rules and excluding the United Nations Convention on Contracts for the International Sale of Goods. Before filing any claim, you agree to first try to resolve the dispute informally by contacting us at support@prefire.trade and allowing thirty (30) days. Any dispute not resolved informally will be subject to the exclusive jurisdiction of the competent courts of the Operator's place of residence, and you consent to venue there. Nothing in this section deprives you, as a consumer, of the protection of the mandatory provisions of the law of your country of habitual residence. To the extent permitted by law, you and the Operator agree that disputes will be resolved on an individual basis and not as part of a class or representative action.
These Terms, together with the Privacy Policy and Refund & Cancellation Policy, are the entire agreement between you and the Operator regarding Prefire and supersede prior agreements. If any provision is held unenforceable, the rest remain in effect and the unenforceable provision is limited to the minimum extent necessary. Our failure to enforce a provision is not a waiver. You may not assign these Terms; we may assign them to a successor. We are not liable for failure or delay caused by events beyond our reasonable control (force majeure). Notices to you may be given by email or by posting on the website; notices to us must be sent to support@prefire.trade.
Questions about these Terms? Contact support@prefire.trade or reach us on our Discord (Prefire).
This Privacy Policy explains how Prefire (the "Operator", "we", "us") collects, uses, and protects personal data when you use Prefire and the https://prefire.trade website. For the purposes of the EU/UK GDPR, the Operator is the data controller. Contact: support@prefire.trade.
We rely on: performance of a contract (to provide the service and billing you request); legitimate interests (security, fraud prevention, service improvement, direct communications about the service, and measuring which of our own marketing campaigns brought you to us — we have weighed this against your rights and consider it low-impact, as it involves no third party, no profiling, and no automated decision about you); consent where required — including for all third-party advertising measurement (Section 7), which never operates without it and which you may withdraw at any time; and compliance with legal obligations (such as tax and accounting records).
We do not sell your personal data. We share it only with service providers who process it on our behalf under appropriate safeguards, including: Stripe (payment processing), our email delivery provider (transactional email), and our hosting/infrastructure provider. Only where you have consented to advertising measurement (Section 7) do we also share advertising identifiers and a hashed form of your email address with the advertising platforms we use to measure our campaigns, which may include Meta Platforms Ireland and X Internet Unlimited (Ireland); if you decline, nothing is shared with them at all. We may disclose data where required by law, to protect our rights, or in connection with a business transfer. When you direct Prefire to act on CSFloat, information necessarily flows to CSFloat as a result of your own instruction; CSFloat's handling of your data is governed by its own privacy policy.
We keep account and licence data while your account is active and as needed to provide the service. Auto-Buy activity history and aggregated usage metrics are retained on a rolling basis (approximately 30 days) unless a longer period is needed for security or legal reasons. Billing and tax records are kept as long as required by law. When data is no longer needed, we delete or anonymise it.
Depending on where you live (including under the EU/UK GDPR and, for California residents, the CCPA/CPRA), you may have the right to access, correct, delete, or receive a copy of your personal data, to object to or restrict certain processing, and to withdraw consent. You may also lodge a complaint with your local data protection authority. We do not sell personal data. Under the CPRA, disclosing identifiers to an advertising platform can count as "sharing" for cross-context behavioural advertising: that happens only where you have consented to advertising measurement (Section 7), and withdrawing that consent — from the cookie settings control on this site, or by emailing us — is also how you exercise the right to opt out of it. To exercise any right, email support@prefire.trade; we will respond within the time required by applicable law and may need to verify your identity.
Strictly necessary. A security/bot-clearance cookie that confirms your browser has passed our verification check, and a session/preference token stored in your browser. These are what make the site work, they are set on the strictly-necessary basis, and they cannot be switched off. You can block cookies in your browser, but parts of the site will then not function.
Advertising measurement — consent only. If you accept, we load measurement tools from the advertising platforms named in Section 4. They set their own cookies (for example Meta's "_fbp", which lasts up to 90 days), read the advertising click identifier in the link you arrived from, and receive a cryptographically hashed version of your email address, so that we can tell which advertisement led to a sign-up or a purchase. Nothing in this category loads, and no such cookie is set, until you accept it — declining is one click and is exactly as easy as accepting. You can withdraw consent at any time from the cookie settings control on this site or by emailing support@prefire.trade, and we will not ask again for at least six months after you decline.
We use technical and organisational measures appropriate to the risk, including password hashing, encryption in transit (HTTPS), access controls, and rate-limiting. No method of transmission or storage is completely secure, and we cannot guarantee absolute security.
Your data may be processed in countries other than your own, including where our providers (such as Stripe) operate. Where required, we rely on appropriate safeguards such as the European Commission's Standard Contractual Clauses.
Prefire is not intended for anyone under 18, and we do not knowingly collect data from children. If you believe a child has provided us data, contact us and we will delete it.
The site links to third parties such as CSFloat, Steam, and Discord. We are not responsible for their content or privacy practices; review their policies separately.
The Prefire CSFloat Bridge is a browser extension published by us. It exists because CSFloat shows some data only to a signed-in browser and blocks server addresses, so the extension carries out Prefire's CSFloat requests in your own browser session, over your own connection.
What it carries is what you have told Prefire to do. Reading the competing buy orders on an item is read-only. If you use the browser app and you arm Prefire to trade, the requests that create, change or cancel your CSFloat buy orders — and that buy a listing with your own money — are performed by this extension, because that is the only route those requests have out to CSFloat. The extension never originates a request of its own and never decides what to buy; it carries only what your own Prefire account asks of it, under the limits you set there.
The desktop application does not use the extension at all — it reaches CSFloat directly, and the relay described above runs only for the browser app, only while a Prefire tab is open. Data the extension passes to Prefire is never sold, never used for advertising, and never used for any purpose other than providing the service you asked for.
We may update this Policy; the "last updated" date above reflects the current version, and material changes will be notified by email or an on-site notice. Continued use after a change takes effect means you accept the updated Policy.
Privacy questions or requests? Contact support@prefire.trade.
This Refund & Cancellation Policy is part of, and should be read with, our Terms of Service. It applies to subscriptions and credit packs, and sets out your responsibility for automated purchases.
You can cancel at any time from your panel or by emailing support@prefire.trade. Cancellation stops the next renewal; your plan stays active until the end of the current paid period, and you are not charged again after that. We do not automatically refund the remainder of a period you have already started.
Prefire is a digital service that gives you access to functionality immediately, so subscription fees are generally non-refundable, including for partially used periods, except where a refund is required by applicable law or granted by us at our discretion as a goodwill gesture. If you believe you were charged in error or the service was materially unavailable, contact us within 14 days and we will review it in good faith.
EU/UK consumers. You normally have a 14-day right to withdraw from a purchase of digital services. By subscribing and starting to use Prefire immediately, you request that the service begin during the withdrawal period and acknowledge that you lose the right to withdraw once the service has been fully performed; where the service is only partly performed, any statutory refund is reduced proportionately. This does not affect your mandatory consumer rights.
Any free plan or free 24-hour trial is provided at no charge and carries no payment obligation. It converts to a paid plan only if you choose to subscribe. Trials may be limited to one per person and may be withdrawn or changed at any time.
Credit packs are one-time purchases of request allowance. Because credits are made available and consumable immediately, they are non-refundable once added to your account, and any unused credit has no cash value and is not redeemable for money.
Auto-Buy and Auto-Bargain act on the rules, targets, and limits you configure and arm, using your own CSFloat account and your own money. You are solely responsible for every purchase, offer, and cancellation that results, including any that stems from a rule, filter, or limit you set, changed, disabled, or overrode.
If you cancel a purchase on CSFloat, CSFloat charges its own cancellation fee (currently 4%). That fee, and the purchase itself, are strictly between you and CSFloat. Prefire does not refund, reimburse, or otherwise cover CSFloat cancellation fees, purchases, or any resulting loss, and does not provide any money-back guarantee for automated purchases. To the maximum extent permitted by law, Prefire is not liable for purchases, offers, cancellations, fees, or losses arising from the automation you configured and armed. If you are not comfortable being responsible for what your rules buy, keep Auto-Buy and Auto-Bargain off.
If you have a billing concern, please contact us first — we can usually resolve it faster than your bank. Initiating a chargeback or payment dispute without first contacting us, or disputing a charge you authorised, is a breach of these terms; we may suspend or terminate your account, contest the dispute with evidence of your use, and recover related costs.
Prices may change as described in our Terms; changes apply to future renewals after notice. Prices may exclude taxes, which are added where applicable.
Need help with a cancellation? Contact support@prefire.trade or reach us on our Discord (Prefire).