no fees on losses is one of the structural choices that defines uponly.win, and we get asked about it constantly. is that real? how does the math work? what is the catch? this is the long-form, honest answer. there is no catch — but the reason there is no catch is more interesting than a marketing line, so we are going to walk through it properly.
the short version: most platforms in this category — perp dexes, casinos, sportsbooks, prediction markets — make money primarily when users lose. uponly inverted that. we only earn on net winnings, and we share half of what we earn with the creator who brought the trader. that one design decision changes everything downstream — the incentives, the support behavior, the product roadmap.
how most platforms actually make money
the truth about how leveraged-trading and gambling-adjacent businesses earn revenue is not a secret, it just rarely gets put plainly. here is the structural breakdown.
- casinos earn through a house edge baked into every game. the math guarantees that, in aggregate, players lose. the casino does not need to root for any specific bettor — the distribution does the work.
- sportsbooks earn through vig (the juice) — the difference between fair odds and the odds offered. over enough bets, the vig compounds into a near-certain margin for the book.
- prediction markets earn through fees on every contract, plus the spread between bid and ask. they earn on volume regardless of which side wins.
- most perp platforms earn through a combination of open fees, close fees, borrow rates, funding spreads, and in many cases the platform itself or its lp pool takes the other side of trader pnl.
none of this is illegal or even necessarily unethical. it is the standard model. but it has a consequence: the platform is structurally rooting for the loss column to grow. user churn through losses is not a bug, it is the revenue engine.
what we did differently
uponly took the same activity — leveraged perp trading — and rebuilt the fee surface from scratch. the rules:
- zero fee to open a position. the platform takes nothing at entry.
- zero fee when a trade loses. when collateral is wiped, the platform earns literally zero from that trade.
- a small variable fee applied only to net winnings when a trade closes with positive pnl. that is the entire revenue model.
- half of every winnings fee flows to the creator who referred the trader. no claw-back, no time limit, no tiering. see the 50 percent creator program.
why we built it that way
three reasons. they compound.
reason one: aligned incentives are an actual moat
if you earn when users lose, your product roadmap quietly bends in that direction. you optimize for churn. you make it easier to add collateral than to withdraw winnings. you push the highest-variance experiences first. you might not mean to do it — the incentives just slowly steer the team.
if you earn only when users win, the optimization vector flips. you want sessions that occasionally produce winners. you want winners to come back. you want winners to bring friends. the product gets sharper around the things that make a winner happen and stickier around the things that keep them around. the same engineering hours produce a different product because the gradient is different.
reason two: it is the right product for the moment
the people who actually want to use a max-leverage perp arcade — meme page audiences, crypto-native degens, sports-bettor adjacent users — are sophisticated enough to read a fee schedule. they have seen the casino model. they have seen the sportsbook model. they have seen the prediction-market fee. "no fee unless you win" is the only message that reframes the entire activity, because everyone understands what every other model is.
reason three: the creator economy needs an honest product to sell
we built uponly as the arcade for meme pages. creators with crypto-adjacent audiences are the distribution engine. for a creator to comfortably push a perp arcade to their audience, the product needs to not feel like the creator is helping the platform extract value from their own followers. "no fees on losses, only on winnings, and half of that fee goes to me as the creator" is a sentence a creator can say in public without flinching. that is the whole point. for the creator monetization context, findclout.com is our growth partner for onboarding meme pages and trading communities at scale.
how the math actually works in practice
two illustrative scenarios.
scenario one — the losing rip. trader posts 50 usdc at 250x leverage. position liquidates two minutes later. trader is down 50 usdc. uponly earns: 0. avantis trading fees apply at the venue layer (those are not ours) — that is the entirety of the cost flow. there is no platform fee, no spread mark-up, no open fee, no close fee from uponly.
scenario two — the winning rip. trader posts 50 usdc at 250x leverage. position closes up 4x to 200 usdc. trader pockets the win. uponly applies a small variable fee to the net winnings only — not the collateral, not the notional, just the profit. of that fee, half flows back to the creator who referred the trader. the trader walks away significantly up, the creator earns ongoing revenue, the platform earns a small cut.
why we can run this model at all
people ask if "no fees on losses" is sustainable. the answer is: yes, because of how the math works in aggregate, not because we are subsidizing anything.
- we are a frontend, not an exchange. settlement happens on avantis on base. we do not pay liquidity costs, market-making costs, or order-book infrastructure costs.
- we have minimal operational overhead. the entire first version of uponly was built in a single night — see uponly was built in one night for the build story.
- winning rips, while rarer than losing rips, generate enough fee revenue per occurrence to fund the operation. one fat win pays for a long tail of churn at zero cost.
- because we earn only on wins, our users are incentivized to actually return after wins. that retention compounds — and retention is the only metric that matters for a session-based product.
the 50 percent creator share, briefly
half of every winnings fee flows to the creator who referred the trader. forever. no claw-back. no time limit. no tiering. no "promotional rate that drops next quarter." this is the most aggressive revenue share in the category by a wide margin, and it is permanent.
the math: if we did not believe the model was sustainable, we could not pay half of it out forever. the fact that we do pay half out forever is the strongest signal we can give that the structural design works.
what this means for you
if you are a trader: every losing rip is free of platform fees. you keep your full attention on the trade, not on whether the venue is nickel-and-diming you on entry and exit. winning rips pay a small fee. that is the whole deal. if you want the broader safety picture, see is uponly.win actually safe.
if you are a creator: the audience you bring earns you 50 percent of their winnings fees, forever. losing trades from your audience cost you nothing and cost them nothing in platform fees either, which means you are never in the awkward position of having referred someone to a platform that grinds them down. the 50 percent revenue share program is the same structural decision applied to distribution.
tap the button (and a CTA)
no fees on losses is the structural choice that defines uponly. it is not a slogan and it is not subsidized — it is the fee logic. if you want to feel a real max-leverage perp where losing trades are platform-fee-free and winning trades pay a small cut, try uponly. perps at 75x to 500x are high-risk entertainment regardless of fee design — size to entertainment, not to your rent.