What is a bonding curve? How launchpad pricing actually works

A bonding curve is a formula that sets a token's price from how much of its supply has been bought. There is no order book and no counterparty: the contract itself sells to buyers and buys from sellers, raising the price as supply is taken and lowering it as supply comes back. It lets a token trade from its first block with nobody having seeded a pool.

The formula, and what "virtual" means

Most memecoin bonding curves are the same constant-product formula that powers automated market makers: two reserves, x and y, whose product stays constant. Buy some of one and you must add proportionally more of the other, and the price you pay is the ratio between them at that moment.

The trick that makes it work at launch is that the reserves are virtual. A normal liquidity pool needs somebody to deposit both assets before anyone can trade. A bonding curve launchpad instead initialises the contract with a notional starting balance of the quote asset — SOL, ETH or BNB — and the token's full supply, and prices trades against those numbers as if the balance were there. Real money accumulates as people buy; the virtual component is what stops the very first purchase from costing infinity.

pump.fun runs this model on Solana with a total trading fee of 1.25% split between the token's creator and the protocol. The exact virtual reserve constants and the current graduation threshold are no longer published in pump.fun's official documentation, and we do not print figures for them here. Anyone quoting precise constants for 2026 is citing an old version of the docs. The mechanism is public; the parameters are not.

Why early buyers get a better price

This is the part that decides who profits, and the maths is unavoidable rather than unfair.

Take an illustration with round numbers that belong to no real platform. Say a curve starts with 1,000 units of virtual quote asset and 1,000,000,000 tokens, so k is fixed at their product. The first buyer spending 10 units receives roughly 9.9 million tokens. A later buyer spending the same 10 units, after 500 units have already gone in, receives around 4.4 million — less than half, for identical money. By the time 2,000 units have accumulated, the same spend buys about 2.5 million.

Every purchase moves the reserve ratio and every subsequent buyer pays more. The curve is not a prediction that price will rise; it is a mechanical consequence of the formula. This is exactly why the first transaction after a deployment is so contested, and why automated buyers win it — see sniping and why you probably lose to bots.

The same curvature works against you on the way out. Selling into a thin curve moves the price down as hard as buying moved it up, which is why slippage settings matter more on a curve than almost anywhere else. What slippage is and how much to set covers the practical side.

What the curve is actually for

Three problems get solved at once, and it is worth being clear about which.

It removes the need for seed liquidity. A deployer with no capital can open a market. That is genuinely democratising and it is also why 11.9 million tokens were launched on pump.fun between January 2024 and June 2026.

It removes the need for a counterparty. Every buy and every sell has a guaranteed fill against the contract at the formula price. There is no waiting for someone to take the other side.

And it defers the moment liquidity becomes real. Until a token graduates, the "liquidity" is a set of numbers in a contract, not a pool anyone can withdraw from. This is a safety property, not a flaw — nobody can pull liquidity that does not yet exist as a position.

Graduation, and what it does to the curve

At a threshold, the curve is retired. The accumulated quote asset and the remaining tokens are moved into a real pool on a decentralised exchange, and from that point pricing comes from the pool rather than the formula.

pump.fun migrates atomically to its own PumpSwap venue, with graduation costing roughly 0.015 SOL and the graduation market cap reported in the $69K–$100K range. hood.fun on Robinhood Chain migrates to a locked Uniswap v3 pool. Four.meme on BNB Chain migrates to PancakeSwap. nad.fun on Monad graduates at approximately 225,000 MON collected and around 80% of supply sold, though its fee parameters and migration target are not stated in reachable documentation and we do not guess at them.

Two consequences for a holder. Your tokens do not move and your balance does not change — the migration happens to the liquidity, not to you. But the price mechanism changes underneath your position, and there is usually a discontinuity at the handover as the curve price and the new pool price reconcile. That transition is covered in detail in what happens when a token migrates to a DEX, and what the threshold actually signifies in what graduation means on a launchpad.

Reaching a DEX is the exception rather than the rule. The defensible range for the 2026 graduation rate is 0.5% to 2% depending on the measurement window, and a widely cited study of 832,941 mints between 8 May and 10 June 2026 put a lower bound at 0.198%, explicitly measured over about six minutes per mint.

The venues that use no curve at all

Not every launchpad has one, and the exceptions are structurally interesting.

Pons.family on Robinhood Chain has no bonding curve and no migration. A fixed 1B supply is paired with a liquidity pool from the first block, and buys and sells route through that same pool forever. Its 4.2 ETH "graduation" is a status flag, not a mechanical event — nothing moves. How Pons works without a curve goes through the trade-offs.

Clanker on Base deploys straight into a Uniswap v4 pool, with no curve phase at all. Pools.trade offers both: an instant-launch mode that is a classic bonding curve, and a Crowd Launch mode that runs a four-hour window of TWAP bids with a $10K minimum FDV, refunding participants if the threshold is not met. Heaven, now defunct, seeded pools with "virtual liquidity" rather than running a curve.

The practical difference for a buyer is that a curveless launch has a real, withdrawable-or-locked pool from the beginning, so the liquidity pool mechanics apply from block one rather than after a graduation event.

What a bonding curve does not protect you from

It does not stop the deployer buying their own supply first at the cheapest point on the curve, which is the normal case rather than the exception. It does not stop coordinated bundled buying across many wallets in the same block. It does not make the token worth anything.

And it does not guarantee an exit. A guaranteed fill at the formula price is not a guarantee of a good price — sell into a curve that has retraced and the formula pays you exactly what the depleted reserve ratio says it should, which can be a small fraction of what you paid.

Finally, the parameters are not constants of nature. Venues change thresholds, fee splits and curve shapes, sometimes without announcement, and pump.fun has stopped publishing several of them. Check the venue's own documentation on the day you act.

Frequently asked questions

How does a bonding curve set the price?

By formula rather than by market. Most memecoin curves use a constant-product relationship between a virtual quote reserve and the token supply: each purchase removes tokens and adds quote asset, shifting the ratio and raising the price for the next buyer. The contract quotes and fills every trade itself, so no counterparty is needed.

Why do early buyers pay less on a bonding curve?

Because price is a function of how much supply has already been bought. On a constant-product curve the same spend buys progressively fewer tokens as the reserve ratio shifts. This is arithmetic built into the formula, not a prediction about demand, and it is why the first block after deployment attracts automated buyers.

What happens to a bonding curve when a token graduates?

It is retired. The accumulated quote asset and remaining tokens move into a real pool on a decentralised exchange — PumpSwap for pump.fun, PancakeSwap for Four.meme, a locked Uniswap v3 pool for hood.fun — and pricing comes from that pool afterwards. Your holdings are unaffected, but the price mechanism changes and there is usually a discontinuity at the handover.

Do all memecoin launchpads use bonding curves?

No. Pons.family and Clanker deploy directly into a liquidity pool with no curve phase, and Pools.trade offers a curveless Crowd Launch mode alongside an instant bonding-curve mode. Curveless venues have a real pool from the first block; curve venues do not have one until graduation.

---

Once the curve is gone, the pool is what buyers check

A bonding curve cannot be rugged because there is nothing to withdraw. A real pool can be, which makes the migration the moment the question changes from "how does the price work" to "who controls the liquidity." Team Finance — built by TrustSwap, which also builds Meme Central — locks LP tokens for a fixed term on Ethereum, Robinhood Chain, Polygon, Base and BNB, and the lock shows as a verified badge on the token's page in the Meme Central feed. A lock secures the pool, not the deployer's own bag, and it does nothing about the price the curve left you at.

---

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.