The launchpad wars: how memecoin launches changed, 2024 to 2026

Memecoin launching went from one venue to a market and back to concentration twice in thirty months. pump.fun opened in January 2024 and still held 85.9% of Solana token issuance by 8 February 2026. What actually changed the board was a new chain: Robinhood Chain went live on 1 July 2026 and ran through three dominant launchpads in nine weeks.

This is the history, with the numbers and their dates attached. The reason it is worth reading rather than a list of names is that each phase settled a specific design question — who pays, who captures the fee, and whether liquidity can be removed — and the answers stuck even where the venues did not.

2024: one venue, and a category that barely existed

pump.fun launched in January 2024 with a mechanism that has since been copied everywhere: a constant-product bonding curve on virtual reserves, zero creation fee, a flat trading fee, and automatic migration of the whole liquidity position to a decentralised exchange once the curve fills. The full mechanism is in what is pump.fun. The important part commercially is that the creation fee was zero. Making launches free converted token issuance from a project decision into a consumable, and volume followed accordingly.

The scale of the category at that point was negligible. CoinGecko's series puts launchpad memecoins at 1.5% of total memecoin market capitalisation in July 2024. Almost all of the category's value was in tokens that predated launchpads entirely.

By 10 June 2026, pump.fun had processed roughly 11.9 million cumulative launches since January 2024, running about 42,000 a day. Across that entire population, 18 tokens had ever exceeded a $10M market capitalisation and 96 had ever exceeded $1M. That ratio is the backdrop to everything below, and it is examined properly in why most memecoins go to zero.

Early 2025: the share peak

Launchpad memecoins peaked at 20.5% of memecoin market capitalisation in January 2025, up from 1.5% six months earlier — a fourteen-fold increase in share in half a year. pump.fun set its all-time single-day fee record of $15.5M in the same month, around its first anniversary.

That January peak has not been retaken. By November 2025 the category was running roughly $1.2bn in daily volume, and the sector-wide contraction visible in memecoin prices — the category fell from a $150.6bn peak in December 2024 to $47.2bn by November 2025 — applied to launchpads as well. Everything after January 2025 is a story of share moving between venues inside a smaller pie, not of the pie growing.

Mid-2025 to late 2025: the first real challenger, and the largest reversal in the space

LetsBonk reached approximately 54% of Solana launchpad share in July 2025. It was the only time pump.fun's dominance was seriously broken, and it did not last.

What LetsBonk looks like now, from its own official revenue dashboard as of August 2026: monthly revenue $84,910, daily average $2,740, 24-hour volume $5,410, 24-hour fees $380.76. From roughly half the chain to a few thousand dollars a day. The mechanical tell arrived before the numbers did — on 3 December 2025 the revenue split was restructured to remove the Buy/Burn and SBR components entirely, replacing them with a "Buy for BNKK" allocation at 51%. A launchpad removing its own flagship burn mechanism is a statement about expected future revenue, and it was made ten months before the fee figures made it obvious. The full account is in what happened to LetsBonk.

Two other venues failed differently in the same window, and the differences are instructive.

Heaven ran no bonding curve at all, seeding roughly 35 SOL of virtual liquidity per pool, and routed 100% of protocol revenue into buying and burning its own LIGHT token — the "God Flywheel." LIGHT went from about $15M to $130M and back to $33M in market capitalisation. The quarterly fee trend tells the rest: $4.48M in Q3 2025, $120K in Q4 2025, $24.8K in Q1 2026, $7.3K in Q2 2026, and $0 across Q3 2026. A reflexive flywheel run in reverse is not a slow decline; it is a step function.

Believe failed a third way, through operations rather than mechanism. It paused on-chain payouts and asked creators to move to PayPal, and its founder faced litigation over the token's collapse. DefiLlama has no protocol page for it, which for a venue that had real volume in 2025 is diagnostic. We are not going to state that it is or is not currently operating, because we could not confirm either as of 31 August 2026. All three cases are set out together in dead and dying memecoin launchpads.

What Base tried, and what its own distribution partner said about it

Base ran the most ambitious non-Solana experiment of the period: creator coins, where the launchpad thesis was social rather than speculative. Zora reached 1.6 million creator coins minted and around $470M in volume by August 2025.

It ended explicitly. Base discontinued Creator Rewards and removed the social feed in February 2026, and on 13 July 2026 Brian Armstrong said of the effort: "It didn't work. We pivoted earlier this year. We messed up, time to move on." The ZORA token fell about 95% from roughly $550M in August 2025 to around $30M. DefiLlama has Zora at $13,775 in 30-day fees as of 31 August 2026. The postmortem is in what happened to Zora.

Base did not empty out entirely. Clanker — an agent that deploys tokens directly to Uniswap v4 pools with no bonding curve, taking a fixed 20% of the creator's LP fee — recorded $234,041 in 30-day fees on 31 August 2026. That is a functioning product at a small scale, which is a different outcome from Zora's and worth separating.

BNB Chain settled into the same shape. Four.meme reported $388,324 in 30-day fees and $118.15M in 30-day DEX volume on 31 August 2026, against cumulative DEX volume of $9.905B — a live venue running well below its own history.

July 2026: a new chain resets the board in nine weeks

The most compressed competitive sequence of the whole period happened after Robinhood Chain's mainnet launch on 1 July 2026, and it is the clearest natural experiment available on how launchpad markets form.

Noxa held the first eleven days. It ran roughly 60,000 launches in under two weeks — about 4,300 a day, around 75% of all chain deployments — and took in the region of $12M in fees. It halted new launches on 11 July citing bot spam, routed 100% of trading fees to creators, and its site went dark on 13 July. It issues nothing now. There has been no conclusive public finding of fraud and we do not describe it as one.

Pons took the next three weeks with a mechanism that inverted the standard design: no bonding curve and no migration at all. Fixed 1B supply, a real liquidity pool live from block one, buys and sells in the same pool forever, and "graduation" at 4.2 ETH of paired WETH as a status flag rather than a mechanical event. Fees are 1% on the pool plus a 0.0005 ETH launch fee, split 70% creator and 30% protocol. At its 27 July peak it processed 1,651,979 trades in 24 hours — about 54% of every transaction on the chain — on $116.8M of volume and 12,384 tokens created. See what is Pons.family. One practical warning that belongs with any mention of it: several lookalike domains rank in search, and only ponsfamily.com is confirmed by the official documentation.

Pools.trade, from Uniswap Labs, arrived on 5 August 2026 and took roughly half the chain's launchpad volume within days on a model nobody else was offering: zero launchpad fee, a 0.25% LP fee that autocompounds into permanently locked liquidity, and every pool a standard Uniswap v4 pool whose liquidity the creator cannot remove. Day one produced about $73.6M of Uniswap v4 volume on Robinhood Chain — more than Uniswap v4 did on Ethereum mainnet the same day, at roughly $47.2M — across around 6,000 tokens. Within days it held about 50% of launchpad volume and 40% of new tokens. The mechanism is covered in what is Pools.trade.

The current standings show why "market share" needs a stated metric. As of 31 August 2026, Pons led all launchpads on seven-day fees at $16.13M, ahead of pump.fun's $14.3M — but Pons's seven-day revenue was only $2.84M, because most of its take passes through to creators. On revenue, pump.fun leads comfortably. On issuance share, pump.fun still holds the Solana market. Three metrics, three different leaders, all true at once. Why the numbers diverge is a fee-design question, set out in how launchpad fee splits actually work.

Where it stands, and what our own index adds

Third-party aggregate figures for 31 August 2026: DefiLlama tracks 245 launchpad protocols holding roughly $240.4M in aggregate TVL, with the top five by seven-day fees being Pons at $16.13M, pump.fun at $14.3M, Flap.sh at $2.68M, Bags at $1.51M and o1 Launchpad at $739K. Across launchpads and trading terminals together, roughly 90% of the visible fee market sits in four names. The monthly reading of that table lives at state of the launchpads.

Meme Central indexes launches across venues on Solana, Robinhood Chain, Base, BNB and Monad, which lets us show the shape of the shift rather than only its endpoints. Everything in the two blocks below is a sample of tokens Meme Central has indexed. It is not whole-market data, coverage begins at a different date per venue, and it should not be summed into a market total.

The coverage envelope is returned with every response from our public API, and the live per-venue view is on the analytics hub. Where our sample and a third-party aggregate disagree, we publish both and say which is which rather than reconciling them silently.

The four questions the wars actually settled

Strip out the names and the same four design questions recur, and the market has now given fairly clear answers to three of them.

Who pays to launch? Nobody, increasingly. pump.fun set the zero-creation-fee norm in 2024; Pools.trade took it further in 2026 by charging no launchpad fee at all and earning from the pool instead. Charging for launches is now a competitive disadvantage.

Who captures the trading fee? The direction is toward creators. Pons routes 70% to creators, Bags lets creators take 1% of every trade split across as many as 100 recipients, and Pons's fee-versus-revenue gap on the league table is the visible consequence. Venues that capture most of what they charge are the exception.

Can the creator remove liquidity? The answer that won is no. Pools.trade locks liquidity permanently at the protocol level, hood.fun auto-migrates into a locked Uniswap v3 position, and pump.fun's atomic full-LP migration leaves nothing to pull. The mechanism that made rugs trivial in 2023 has been designed out of the major venues, which is a genuine improvement and addresses exactly one of the many ways a token fails.

Does a bonding curve need to exist at all? This is the open one. Pons and Clanker say no and are both running. pump.fun says yes and holds the largest issuance share. There is no verdict yet, and anyone offering one is guessing.

What this history does not tell you

It does not predict the next phase. Robinhood Chain reorganised three times in nine weeks; a timeline is a record, not a forecast, and we make no claim about which venues persist.

It is not complete. Roughly fifteen launchpads competed on Robinhood Chain alone, most of which never got enough coverage to be documentable. Telegram-native venues are poorly indexed by aggregators, so their absence from fee tables is partly a measurement gap rather than evidence of death — that is an inference and we present it as one.

The metrics do not agree with each other, and that is not a flaw in the reporting. Fees measure extraction, volume measures activity, issuance measures launch counts, and revenue measures what a venue keeps. All four have pointed at different winners during 2026. Any single-number ranking has silently chosen one.

And nothing here says a launch on any of these venues is likely to work. The survival data says the opposite. A side-by-side of what each venue does mechanically, rather than where it sits on a league table, is in memecoin launchpads compared.

Frequently asked questions

Which memecoin launchpad has the largest market share?

Name the metric first. On seven-day fees as of 31 August 2026, Pons led at $16.13M against pump.fun's $14.3M. On protocol revenue, pump.fun leads comfortably because Pons passes most of its take to creators. On Solana token issuance, pump.fun held 85.9% as of 8 February 2026. All three are current and they disagree.

When did memecoin launchpads become significant?

CoinGecko's series puts launchpad memecoins at 1.5% of memecoin market capitalisation in July 2024, peaking at 20.5% in January 2025. That January peak has not been retaken. By November 2025 the category ran roughly $1.2bn in daily volume, inside a memecoin market that had itself fallen from $150.6bn in December 2024 to $47.2bn.

Why did LetsBonk collapse?

It went from roughly 54% of Solana launchpad share in July 2025 to $2,740 in average daily revenue by August 2026, on its own dashboard. No single event explains it. The clearest internal signal came on 3 December 2025, when the Buy/Burn and SBR components were removed from its revenue split entirely — a venue retiring its own burn mechanism.

Did any launchpad mechanism actually improve safety?

One did. Permanently locked or non-removable liquidity is now standard at the largest venues: Pools.trade locks it at the protocol level, hood.fun migrates into a locked position, and pump.fun's full-LP migration leaves nothing to withdraw. That closes the classic liquidity-pull rug. It does not stop creators selling their own allocation, and it does nothing about supply concentration.

Is Meme Central's launch data the whole market?

No. Our figures cover tokens Meme Central has indexed, with a coverage start date that varies by venue, and every API response carries that envelope. Third-party aggregates on this page are labelled with their source and reading date, and the two are kept visibly separate rather than blended into one series.

---

If you are launching, the fee model is not the decision that matters

Every venue in this history competed on fees and mechanics, and the tokens that failed did so for reasons no fee schedule touched. MintPlus — from TrustSwap, which also builds Meme Central — takes the opposite approach to a bonding curve: a fixed supply you define, deployed with liquidity locked through Team Finance at creation, on Ethereum, Robinhood Chain, Polygon, Base and BNB. It gives you no bonding curve, no built-in distribution and no audience, which is most of what a launchpad actually sells. If what you need is a crowd on day one, a launchpad is the right tool and this is not it.

---

Nothing here is financial, legal or tax advice. Memecoins are extremely high-risk: most lose most of their value, and the majority of tokens launched never reach a decentralised exchange at all. Never spend money you cannot afford to lose entirely. Meme Central does not recommend any specific token. Data described as Meme Central's own reflects tokens indexed by Meme Central and is not whole-market data.