The short answer
Pump.fun and Fomo are fighting over the same customer: the person who wants to speculate right now, on their phone. Pump.fun owns the machinery underneath almost every Solana memecoin trade. Fomo owns the app people actually enjoy using. In August 2026 that arrangement turned into a war.
- Pump built the factory. Since January 2024 it has minted close to 12 million tokens and booked north of $800 million in revenue, one of the most profitable products crypto has shipped.
- Fomo built the room. Founded in May 2025 by three ex-dYdX engineers, it turned trading into a feed you can follow, and raised $75 million at a $550 million valuation in June 2026 on the strength of it.
- Partners underneath, rivals on top. Fomo's trades often settle on infrastructure Pump owns. Pump now wants the customer relationship too, and is shipping social features to take it.
- The money is asymmetric. Pump gives its front end away free because bonding curves and PumpSwap already pay the bills. Fomo charges because the experience is the entire business.
- The psychology is not new. Peer-reviewed gambling research has described this exact product for years. The label it uses is gamblification.
In January 2024 a website went live that let anyone create a cryptocurrency in about thirty seconds, for roughly the price of a coffee. No code. No lawyers. No liquidity to put up. You typed a name, uploaded a picture, clicked a button, and a token existed. Most were jokes. A lot were worse than jokes. Within eighteen months that website had become the single most profitable application in crypto.
Eighteen months later, in May 2025, three engineers who had just left the derivatives exchange dYdX started building something that looked nothing like it. Their app hid every piece of blockchain plumbing behind a clean consumer interface. No seed phrase, no gas, no bridges. You signed in with Apple ID and funded with Apple Pay. And what made it spread was not the technology. It was a feed. You could watch other people trade, follow the ones who were good at it, and see it play out in public.
Those two companies are now in an open fight, and what they are fighting over is the revealing part. Not a blockchain or a protocol standard. They are fighting over who owns speculation as a consumer habit. This piece covers where each came from, how they built their audiences, what set the conflict off in August 2026, how their tokenomics differ, and what decades of gambling research says about the product both are racing to perfect.
Section 01
Pump.fun and the problem it actually solved
To understand why Pump.fun got big so fast, you need to see what launching a memecoin looked like before it.
The old way: you deployed a token contract, then created a market by depositing your own money, often thousands of dollars of crypto, into a liquidity pool. That pool was controlled by whoever made it, so the person who launched the token could withdraw everything and vanish. This is a rug pull, and in 2023 it was less an occasional risk than a business model.
Pump.fun, founded by Alon Cohen, Noah Tweedale and Dylan Kerler, killed that specific problem with a piece of maths called a bonding curve.
What a bonding curve actually does
Think of it as a vending machine that sets its own prices. The contract holds the tokens and quotes a price that rises with every purchase and falls with every sale. Nobody seeds a pool because the pool is the contract, and no founder can drain it because no founder controls it. Launching costs about 0.01 SOL. Hit roughly $69,000 in market value and the token graduates, moving liquidity to a public exchange.
That one decision removed the biggest friction point in memecoin creation and the numbers went vertical. Around 1,500 tokens in February 2024. Close to two million by the end of that year. Roughly 11.9 million lifetime from more than 22 million wallets, with cumulative revenue past $800 million. It became the fastest application in crypto history to reach $100 million in revenue, and at points in 2025 accounted for as much as 71% of all Solana token launches.
Here is the number that matters more. Of roughly 12.4 million tokens launched, about 144,000 have ever graduated. A bit over one percent. Analytics firm Solidus Labs found that approximately 98.6% of tokens on the platform displayed rug-pull behaviour. The bonding curve stopped founders draining the pool. It did nothing to stop them dumping their own bags on buyers thirty seconds after launch.
The token, and the money
On 12 July 2025 Pump.fun sold its own token. $PUMP has a fixed supply of one trillion, 33% allocated to the sale. The public portion cleared $600 million in under twelve minutes at a $4 billion fully diluted valuation, roughly three quarters of it executing on-chain through Pump's own interface. Counting the private round, reported totals reach about $1.3 billion.
Buybacks hold it up. A share of protocol revenue is programmatically routed into repurchasing $PUMP on the open market. The exact share is reported inconsistently, from a quarter to all daily revenue depending on the source, so treat any single figure with caution. The direction is consistent: revenue in, token bought back.
It has not been smooth. $PUMP fell close to 70% in the fortnight after the sale, peaked around $0.0088 in September 2025, then gave back three quarters of that. By August 2026 it traded near $0.002 at a $1.2 billion market cap, having run up 87% in thirty days on the back of the fight below.
The platform also carries real legal weight. Class actions filed in early 2025 in the Southern District of New York allege that memecoins issued through it are unregistered securities. There is a UK regulatory ban. And livestreaming was pulled in November 2024 after a thirteen-year-old promoted his own token on camera, dumped it at a $1 million market cap and walked away with around $50,000.
How two companies ended up in the same room
Figure 1. The sequence that turned an infrastructure relationship into a competitive one.
Section 02
Fomo, and the bet that crypto had a taste problem
Paul Erlanger and Se Yong Park left dYdX in 2025 with a thesis that sounds obvious now and did not at the time. Crypto's problem was never the technology. It was that using it felt like homework.
They recruited Prashan Dharmasena as a third co-founder, drew up a list of about 200 people they wanted on the cap table and got roughly 140 in, including Solana co-founder Raj Gokal and Balaji Srinivasan. The name, for what it is worth, is short for fearless. Not the acronym everyone assumes.
The product removed things rather than adding them. No seed phrase, no gas token, no bridges. One dollar balance working across Solana, Base, BNB Chain and Monad at once, each trade settling on the relevant chain while the user sees none of it.
Fomo also never touches customer money. Key generation is split, so the company cannot unilaterally move funds and users can export keys at any time. That sounded philosophical in 2025. In March 2026 it became a legal advantage, when SEC and CFTC guidance extended broker-dealer exemptions to non-custodial interface providers, the exact category Fomo occupies.
The feed is the product
All of that is table stakes though. What made Fomo spread is that it made trading legible to other people.
Positions are public. You follow a trader, watch what they are buying close to real time, check their history, tail them or take the other side. This is not copy trading in the old automated sense: the app does not fire orders on your behalf. You watch, then you decide, which is a meaningfully different psychological experience and one worth coming back to.
Traction came fast. More than 120,000 users in six months, daily revenue near $150,000 by the end of 2025. By mid-2026: over 600,000 signups, roughly 30% of whom had ever placed a trade, and about 70,000 arriving through fiat on-ramps rather than crypto rails. That last number is the one that matters. Those are people who were not already in crypto.
In June 2026 Index Ventures led a $75 million Series B at a $550 million valuation, alongside Union Square Ventures and Benchmark, with an angel list that read like a consumer tech reunion: Mark Pincus of Zynga, Discord chief executive Humam Sakhnini, Eventbrite's Kevin Hartz. Cumulative volume had passed $4 billion. Index matters because they are not a crypto specialist, and generalist funds have mostly avoided consumer crypto trading since FTX collapsed in 2022.
Section 03
How each one recruited its crowd
They grew through different doors, and that explains most of what followed.
Pump grew through supply. Every token came with a creator who had a financial reason to promote it, handing Pump tens of thousands of unpaid marketers pushing links into Telegram groups and X replies daily. The brand became a verb in the trenches. The culture is loud, fast, self-aware about how ridiculous it all is and completely unsentimental.
Fomo grew through people. It only works if there are traders worth following, so its growth loop runs through individuals with audiences. A trader brings followers, followers watch, then trade, then some build followings of their own. Closer to how a social network grows than an exchange, and it produces a different user: calmer, newer, likelier to have arrived from a debit card than a wallet.
Both market speculation, just dressed differently. Pump sells the casino floor at three in the morning. Fomo sells the group chat with people who seem to know what they are doing.
Section 04
The week it turned into a fight
For most of 2025 and early 2026 the relationship was symbiotic. Fomo built the interface people liked. A large share of the trades it produced settled on infrastructure Pump owned. Fomo charged for the experience, Pump took its cut underneath, everyone ate.
Then Fomo got big enough to be a threat. Through July and into August 2026 it logged six consecutive weeks of record volume, with Solana accounting for more than half its flow. Weekly revenue crossed $2 million. On 6 August it briefly passed Axiom in daily fees, which in that corner of the market is a scoreboard moment people notice.
The response landed the next day. On 7 August, Pump.fun shipped a social trading feature of its own and posted record daily active users. Nobody pretended it was a coincidence.
The asymmetry that decides this
Pump also went to zero app fees, and this is the part that decides things. It can give the front end away because it already monetises the layer beneath: bonding curve fees on every launch, then swap fees on PumpSwap. The interface can run at a loss forever because it is a customer acquisition channel for a business that already exists.
Fomo cannot. The experience is the business. Critics of its fees mostly miss what is being paid for, which is not execution but everything around it: discovery, reputation, alerts, the feed, the sense that you are not doing this alone.
There is a more aggressive dimension too. Multiple reports, none confirmed by either company, describe Pump approaching prominent traders with exclusivity contracts worth up to $30,000 a month to move their activity onto Pump and keep it there. Lawyers who have commented publicly say such arrangements are generally legal. They also raise obvious questions about disclosure, given those same people recommend tokens to audiences who assume they are watching independent conviction.
Both companies are building the same thing, and that thing already exists. It is called a casino. The difference is that a casino has to tell you the house edge.The thing nobody in this fight says out loud
Zoom out and the pattern is familiar rather than crypto-specific. Infrastructure that becomes essential eventually climbs into distribution, because distribution is where margins and customer relationships live. Uniswap is doing the same thing right now, moving from plumbing everyone builds on to promoting its own launchpad. Pump owns the pipes of Solana speculation and has decided pipes are not enough.
| Dimension | Pump.fun | Fomo | Who it favours |
|---|---|---|---|
| Founded | January 2024 | May 2025 | Pump on maturity |
| Core business | Launch and swap infrastructure | Consumer interface and social layer | Different games |
| How it earns | Bonding curve plus PumpSwap fees | Fees on the trading experience | Pump can subsidise |
| Token | $PUMP, revenue-funded buybacks | None. Equity and VC funded | Unresolved |
| New-user friction | Wallet required, crypto-native | Apple ID and Apple Pay, no wallet needed | Fomo by a mile |
| Legal exposure | US class actions, UK ban | Fits March 2026 non-custodial exemption | Fomo, for now |
| Main risk | Reputation, regulation, token overhang | Retention, and whether the signal stays honest | Watch both |
From the insider files
Crypto casinos with their own tokens, compared-> SHFL token breakdown: buybacks, burns and what actually drives demand-> NFT marketplaces in gambling->Section 05
Tokenomics, and why only one of them has a token
This is where the businesses diverge most sharply, and the same split runs through crypto gambling.
$PUMP is a revenue-share instrument wearing a memecoin's clothes. Protocol income converts into buy pressure through programmatic repurchases, so holding it is a bet that Solana speculation keeps happening at volume. You are not backing a roadmap, you are underwriting flow. The obvious criticism is that revenue routed into buybacks is revenue not funding development or security work, which matters for a company still facing litigation. The counter is that the ICO left a balance sheet that can afford it.
Crypto casinos have run variations of this playbook for a while, with mixed enough results to be instructive. Our breakdown of SHFL's buyback-and-burn mechanics covers what happens when platform revenue is wired directly into token demand, and the wider comparison of casinos running their own tokens shows how differently it plays out depending on whether the underlying business actually grows.
Fomo has no token and has said nothing about wanting one. It raised roughly $94 million across seed, a $17 million Series A and the $75 million Series B, and is building toward equities, perpetual futures and prediction markets rather than an airdrop. Slower path, cleaner regulatory story, and after the March 2026 guidance it looks like the deliberate choice it probably always was.
Section 06
What the research actually says about this product
Time to stop describing the fight and start describing the thing being fought over, because academic literature has studied it for years and the vocabulary is unflattering.
The mechanism at the centre of memecoin trading is variable ratio reinforcement. Rewards arrive at unpredictable intervals in unpredictable sizes. Sometimes big, sometimes small, mostly nothing. Behavioural psychologists identified this decades ago as the single most powerful schedule for producing behaviour that resists extinction, which is the clinical way of saying it is very hard to stop. It is also, precisely, the reward schedule of a slot machine.
The neurochemistry is more specific than people assume. Dopamine release in the nucleus accumbens peaks during anticipation of an uncertain reward, not on receiving it. The hook is the moment after you buy and before you know. So a product designed to maximise how many of those moments happen per hour is, whatever its intentions, optimising the exact variable that drives compulsive behaviour.
None of this is speculative. A 2026 paper in the Journal of Gambling Studies applies the epidemiological triad used for infectious disease to crypto trading in young adults, splitting risk into host, agent and environment. The host is the individual: impulsivity, sensation seeking, fear of missing out, cognitive bias. The environment is a market that never closes. The agent, in the authors' framing, is the platform, which they describe as implementing a predatory architecture built on gamblification, variable-ratio reinforcement and dark patterns.
Read that back. It is not a description of a casino. It is a description of a trading app.
The numbers behind the label
A few findings worth carrying. A survey of 700 cryptocurrency traders, drawn from a behavioural addiction sample of over 4,000, found high rates of problematic gambling on the Problem Gambling Severity Index. Work published in 2026 found PGSI scores predicted crypto-trading harms better than direct exposure measures like how much someone had invested or how often they checked prices.
Oksanen and colleagues ran a three-wave national survey tracking crypto trading against rising gambling problems over time rather than at a single moment, which matters because cross-sectional work cannot separate cause from correlation. A 2024 study by Weiss-Cohen found higher volatility drove higher trading frequency, holding after controls for financial literacy, age, gender and overconfidence. Volatility itself is the trigger. A 2025 study in European Addiction Research found problematic traders scored higher on urgency and lower on premeditation, acting on distress and thinking less about consequences. And a 2022 paper by Newall and Weiss-Cohen has a title that does the work on its own: The gamblification of investing: how a new generation of investors is being born to lose.
Be precise about the diagnostic position, though. Gambling disorder is the only behavioural addiction formally recognised in the DSM-5, and trading addiction has no separate classification. But clinicians specialising in behavioural addiction routinely assess it using the same criteria, because the mechanisms are close enough that the distinction stops being useful in a treatment room. Our guide to the science of stopping gambling addiction covers onset patterns and what recovery looks like.
The same loop, two industries
Figure 2. Structural comparison drawn from gambling and behavioural addiction literature. Similarity of mechanism does not imply equivalence of intent.
The part the research has not caught up with
Every mechanism above has been studied for years. The social layer has not, and that is what makes this rivalry genuinely new.
Traditional gambling is solitary or anonymous. You sit at the machine by yourself. What Fomo built, and Pump has now copied, adds three things on top of variable ratio reinforcement that the literature has barely examined: social proof (other people are buying this, so it must be reasonable), observational learning (I watched them win, so I can learn to win) and parasocial trust (I have followed this person for months, so I believe them).
Layer those onto a schedule already known to produce compulsive behaviour and you have something research does not yet have a name for. The gambling industry arrived at the same design from the opposite direction: crypto casinos have run community chat, rain drops and streamer integrations for years because social presence keeps people at the table longer. The same convergence shows in how NFT marketplaces get absorbed into gambling products, where tradability turns a prize into a position.
A trust problem sits inside all of it, and it is the thing most likely to decide who wins. Social trading is worth exactly as much as the honesty of the signal. Someone can buy publicly through one wallet and sell through another, build a position quietly before calling it, or dump on the people following them. Blockchains make social trading possible, but transparency of transactions is not transparency of intent. If the feed becomes a place where people get farmed, the premise collapses for whoever owns it.
Section 07
Where this actually lands
Pump probably wins on economics. You cannot beat free, and it can run its consumer app at a permanent loss because the toll booth underneath keeps collecting. Every month fees stay at zero is a month Fomo has to justify its price.
Fomo probably wins the audience that matters most over the next five years. Seventy thousand people arriving through fiat rails is not a rounding error, it is evidence the product reaches people crypto never has. Pump's brand is superb inside the trenches and a liability outside them. Class actions, a UK ban and a 98.6% rug-pull figure do not read well to someone opening their first trading app.
The thing neither can buy is trust, and both are one scandal away from learning how fragile it is.
For anyone reading this from the gambling side, the convergence runs both directions. Casinos have spent five years becoming more like trading platforms: tokens, buybacks, staking, portfolios, charts. Trading platforms have spent the same five years becoming more like casinos: social layers, streaks, leaderboards, near-miss feedback. Somewhere in the middle they meet, and the only meaningful difference left will be that one side has to publish its house edge and file with a regulator.
That is not an argument for regulating memecoins. It is an argument for being honest with yourself about which activity you are doing when you open the app.
Frequently asked questions
What is the difference between Pump.fun and Fomo?
Pump.fun is infrastructure. It lets anyone create a token using a bonding curve and earns fees on launches and on swaps through its own exchange. Fomo is a consumer app that makes trading social, letting users follow other traders and see positions in public. Many trades placed through Fomo settle on infrastructure Pump owns, which is why they were partners before they became rivals.
What is a bonding curve, in plain terms?
It is a pricing formula built into a smart contract. The price of a token rises slightly with every purchase and falls slightly with every sale, and the contract itself holds the liquidity. This removes the need for a founder to deposit their own money to create a market, which also removes their ability to withdraw that money and disappear.
Does Fomo have a token?
No. Fomo has raised approximately $94 million in equity across a seed round, a $17 million Series A led by Benchmark in November 2025, and a $75 million Series B led by Index Ventures in June 2026 at a $550 million valuation. The company has not announced any plans for a token.
Is memecoin trading the same as gambling?
Legally no, structurally very close. Peer-reviewed research finds that high-frequency speculative trading shares the core mechanisms of gambling: variable ratio reinforcement, anticipation-driven dopamine response, near-miss effects and loss chasing. Gambling disorder is the only behavioural addiction recognised in the DSM-5, and trading addiction has no separate classification, but clinicians commonly assess it using the same criteria.
Why can Pump.fun offer zero fees when Fomo cannot?
Because they monetise at different layers. Pump already earns from bonding curve fees on token launches and swap fees on PumpSwap, so its consumer app can run at a loss as a customer acquisition channel. For Fomo, the trading experience is the entire business, so fees on that experience are its revenue rather than a supplement to it.
Is social trading safer than trading alone?
Not necessarily. Public positions make activity visible but not honest. A trader can buy through one wallet and sell through another, build a position before publicly calling it, or exit into the buying pressure of their own followers. Blockchain transparency shows transactions, not intentions, and social proof can increase confidence without increasing accuracy.
Keep reading
This article is informational only and is not financial, investment or gambling advice. Figures reflect reporting available in August 2026 and move quickly, so verify current numbers before acting on them. Reports of exclusivity payments to traders are unconfirmed by the companies involved. Speculative trading and gambling both carry real financial risk and can become compulsive. If either is affecting you or someone you know, contact a support service such as BeGambleAware or GamCare.