The Pump.fun bonding curve, in plain arithmetic

A bonding curve prices a token from a formula instead of from a pool of buyers and sellers. There is no counterparty, no order book, and no liquidity in the usual sense. Understanding that changes how you read early price action and explains why the curve is the most expensive phase to trade against.

Volion Research Updated Jul 30, 2026 4 sections

What a bonding curve actually is

A bonding curve is a formula that sets price as a function of how much supply has been sold. Buying mints tokens and moves price up the curve; selling burns tokens and moves it back down. The contract is always willing to trade at the formula price, so there is no need for a counterparty. Price is therefore fully determined by position on the curve, not by anyone's opinion of value.

The practical consequence is that early price movement on a Pump.fun token carries less information than it appears to. On an ordinary pair, a rising price means buyers outbid sellers. On a curve, a rising price means tokens were bought, and that is all it means. The same purchase produces the same price movement regardless of who made it or why.

It also means there is a hard ceiling on how much volume the curve can absorb: the supply remaining to be sold. A campaign cannot produce arbitrary volume against a curve that is nearly complete, which is a constraint that does not exist on a pool.

Why the shape matters more than the starting price

Bonding curves are typically convex: each successive token costs more than the last, so the price impact of a fixed SOL purchase shrinks as the curve fills. Early buyers move price dramatically with small amounts; later buyers move it far less with the same amount. This is a property of the formula, not a market condition, and it is fully predictable.

For campaign design this is the single most useful fact about curves. The same 1 SOL of buying produces a large price move near the start of the curve and a small one near the end. If your intent is volume rather than price movement, trading against a nearly-complete curve is more efficient because you consume less price impact per SOL of volume produced.

Conversely, if a campaign runs early on a fresh curve, the price effect is pronounced and the SOL cost per unit of volume is higher, because your own buying works against you. That is price impact rather than a fee, so it appears nowhere in a fee table while being one of the largest real costs. Full model in the cost breakdown.

Why curve transactions fail so often

Solana executes transactions in parallel by locking the accounts each one writes to. Every trade against a given curve writes to the same curve account, so all of them must serialise, and transactions that cannot get their lock in the current slot are dropped rather than queued. New launches attract many parties competing for exactly that account, so failure rates run well above the network average.

Measured in our most recent block sample, 92.4% of transactions touching the Pump.fun bonding curve program failed, against a network-wide rate of 27.9%. Every one of those failures paid its fee and produced no volume. Method and per-program tables on the measurement page.

This is a structural property of a single shared account under contention, not a defect in anyone's software. It has two implications worth acting on. First, budget curve campaigns from attempted transactions rather than intended ones, because the gap is large. Second, a fleet trading the same curve in the same slots contends with itself as well as with outsiders, so spreading trades across slots reduces both failures and cost. Mechanism in priority fees.

What graduation changes

When the curve completes, the token migrates to an AMM pool and the pricing mechanism changes entirely. Price becomes a function of pool reserves rather than supply sold, slippage becomes a function of depth, a counterparty exists, and contention typically falls. Any campaign configured for curve conditions needs different settings after migration.

Migration is also the moment new attention tends to arrive, since graduation is itself a visible event. A campaign that stalls during the handoff goes flat exactly when it matters most, which is why automatic handoff is a routing requirement rather than a convenience. Post-graduation configuration is covered in Raydium routing.

One asymmetry worth planning around: the curve has a volume ceiling set by remaining supply, while a pool has one set by depth and your tolerance for price impact. A campaign sized for the curve may be either impossible or trivially small after migration, depending on how deep the resulting pool is.

Running a campaign against the curve rather than reading about it is covered on the Pump.fun volume bot page, and the half that begins at graduation on PumpSwap routing. The venue-by-venue picture for a general Solana volume bot campaign, including Raydium, Meteora DLMM and Orca Whirlpools, is on the overview.

Questions

What is a bonding curve in simple terms?

A formula that sets a token price from how much supply has been sold. Buying mints tokens and pushes price up the curve, selling burns them and pushes it back. The contract always trades at the formula price, so no counterparty is needed.

Why does early price movement look so dramatic?

Because bonding curves are convex: each successive token costs more than the last, so a fixed SOL purchase moves price far more near the start of the curve than near the end. It is a property of the formula rather than a signal about demand.

Why do so many bonding curve transactions fail?

Every trade against a curve writes to the same curve account, and Solana drops transactions that cannot lock the accounts they need in the current slot. New launches attract many competitors for that one account, which pushes failure rates well above the network average.

Is there a limit to how much volume a curve can absorb?

Yes, set by the supply remaining to be sold. This constraint does not exist on a liquidity pool, where the limit is depth and your tolerance for price impact instead.

What happens at graduation?

The token migrates to an AMM pool. Pricing shifts from supply-based formula to reserve-based, slippage becomes a function of depth, and contention usually falls. Campaign settings tuned for the curve should be changed after migration.

Pump.fun volume bot Raydium routing Measured transaction costs Glossary