What Is a Bonding Curve, and Why So Few Tokens Graduate

Most memecoin launchpads do not put your token on a decentralised exchange. They put it in a queue, and the exchange is the prize at the end. The mechanism is the bonding curve; the exit from it is called graduation.

This guide explains both, then publishes the number: how many tokens make it out, where that number comes from, and what it does and does not measure.

What a bonding curve is

A bonding curve is a pricing formula held in a contract. Instead of two assets sitting in a pool and finding a price between them, one contract holds the token supply and a formula says what the next unit costs, based on how much has already been bought.

Buy, and the price moves up the curve. Sell, and it moves back down it. There is no counterparty and no order book, which is why a curve launch needs no starting liquidity and no second asset: the formula supplies the price that a pool would otherwise have to discover.

Two consequences follow immediately. The first is that trading on a curve is not trading on a DEX. Your buy is a transaction with the launchpad's contract, not a swap in a pool, so the token is not routable by aggregators, not indexed on most charting sites, not reachable by trading bots and not available to a limit order. The second is that the curve is owned by the platform, so whatever happens next is the platform's decision, encoded in its contract.

What graduation is

Graduation is the threshold event. When cumulative buying pushes the token past a market-cap or funds-raised line the platform has set, the contract stops being the market. The funds accumulated in the curve are used to open a real liquidity pool on a public DEX, the remaining token supply goes in alongside, and from that moment the token trades like any other pair.

That is the moment a token becomes visible to the rest of the market: a chart on DexScreener or any chart that indexes the chain's pools, routing through aggregators, bots and trackers. Everything most traders think of as "a memecoin trading" begins at graduation, not at launch.

What happens to the LP at graduation

This is the part with lasting consequences, and it varies by platform.

When the pool is opened, someone has to hold the LP position — the claim on the liquidity and on the trading fees it accrues. Most curve launchpads dispose of it rather than hand it over. Burning the LP tokens sends that claim to an address nobody controls, which makes the liquidity permanently unpullable and, in the same stroke, makes the accrued fees permanently unclaimable. Nobody can pull it; nobody can collect from it either.

The alternative is to lock the position rather than destroy it. On Arc, Sharc states that liquidity is "locked forever at graduation, so it can't be pulled" (a lock claim, not a burn; fee treatment not published), read on its own site on 14 September 2026. A lock under a third-party contract for a defined term immobilises the liquidity while leaving the fee stream claimable — the model TrustSwap's Bullcheese uses through Team Finance's audited contracts, per TrustSwap, the operator (Sept 2026; see bullcheese.fun/locked-not-burned). Burned, permanently locked and locked-for-a-term are three different things, and only reading the venue's own documentation tells you which one a token is under.

For a buyer the practical question is not which is better but which is verifiable. A burn is a transaction you can find. A lock is a contract with an expiry you can read. A promise is neither.

The number: how many tokens graduate

Fewer than 2% of all Pump.fun tokens have ever graduated from the platform's bonding curve to a decentralised exchange, according to on-chain data tracked by The Block and reported by Solana Compass on 10 June 2026. The same report notes Dune Analytics dashboards putting the historical rate at roughly 1.4%.

That is the cumulative, all-time figure, and it is the one worth quoting, because the daily rate is a different measurement and a much noisier one. The Block reported on 29 July 2026 that the daily graduation rate reached 6.7% on a single Friday after Pump.fun's BOOST changes altered launch incentives — roughly eight times the June average — against 4.7% across the preceding four days and 2.5% the week before.

So: a cumulative rate under 2%, and daily rates that have ranged from well under 1% to nearly 7% depending on the month and the incentive regime. Every one of those is a small single-digit percentage. None of them is close to a majority.

What the number does not measure

Four limits, all of which matter if you are going to repeat the figure.

It is Pump.fun, not bonding curves in general. Every platform sets its own graduation threshold and its own incentives. A venue with a lower threshold will graduate a higher share of its tokens, and nothing in this data tells you what any other launchpad's rate is.

Cumulative and daily are not interchangeable. The all-time rate is dominated by Pump.fun's enormous back catalogue, including long stretches of very low graduation. A daily figure describes one day's cohort. Quoting a daily spike as if it were the platform's record, or the all-time rate as if it described this week, are both wrong.

Graduating is not succeeding. Reaching a DEX means the token cleared a threshold. It says nothing about whether it still has liquidity, holders or a price a week later.

Academic estimates measure something narrower. A survival analysis of 832,941 Pump.fun launches (revised 13 August 2026) reports a pooled graduation rate of 0.198%, but its authors state plainly that their observation window catches only graduations happening within about six minutes of launch, making every rate in the paper a lower bound rather than a comparable figure. It is good evidence that same-minute graduation is rare. It is not a competing headline number.

Why the rate is so low

Three reasons, in roughly this order of weight.

Launching is close to free, so nearly everything gets launched — tests, jokes, duplicates, batches deployed by scripts. The denominator is enormous by design. Second, the threshold requires real money from strangers, and a token nobody outside the creator's group hears about cannot raise it. Third, the curve is invisible to the wider market until it graduates: no chart to stumble onto, no aggregator routing, no bot flow. The mechanism that makes launching cheap is the same one that keeps almost every launch unseen.

How to verify this yourself

Do not take the number from this page. Take it from the trackers.

The Block's daily chart. "Pump.fun Percent Graduated Tokens (Daily)" on theblock.co plots the daily rate over time. Read the axis carefully — this is the daily series, not the cumulative one.

Dune dashboards. Community-maintained Pump.fun dashboards on Dune publish launch and graduation counts. Check the query date range before quoting anything, and check when the dashboard was last refreshed.

The pool itself. For any individual token, the check that matters is whether a real DEX pool exists. Look the contract up on a block explorer and on a chart that indexes the chain's pools. No pool means the token has not graduated, whatever the launchpad's interface shows.

Why this matters on Arc

On Arc, the field splits. Some venues run a curve with a graduation event, so every token starts life inside the odds above. Others open a real pool on day one, which removes the threshold entirely — a token from those venues is on a DEX from the first block, a statement about tradability and not about quality. Our guide to the Arc launch venues sets out which venue does which, from each one's own published material.

Knowing whether the token in front of you has been through a curve, is still in one, or never had one is the most useful thing you can know about how it will trade.

Bullcheese is a permissionless launch venue. Tokens launched on it are created by anyone, carry no endorsement, and can go to zero. Nothing here is financial advice.

Not financial advice. Memecoins are extremely high risk.

Disclosure: memecentral is built by TrustSwap, and Bullcheese is a TrustSwap product.