Onchain Atlas

Bonding Curve

A token's price is set by a fixed mathematical formula tied to its supply, so buying always mints new tokens at a higher price and selling always burns tokens at a lower price.

Also called: continuous token model · algorithmic market maker · mint/burn curve

What it is

A bonding curve is a smart contract that acts as an automatic, always-available buyer and seller of a token, using a predetermined formula (often linear or exponential) that relates the token's price to how many tokens currently exist. Instead of trading against other people on an order book or a liquidity pool, users trade directly against the contract itself: buying mints new supply and pushes the price up along the curve, selling burns supply and pushes the price down. There's no need for a separate liquidity provider — the curve itself is the market maker.

How it works

  1. A developer chooses a curve formula, for example price = supply² (exponential) or price = k × supply (linear), and deploys it in a contract holding a reserve asset (like ETH or a stablecoin).
  2. A buyer sends reserve currency to the contract; the contract calculates how many new tokens to mint based on the integral of the curve between the current supply and the new supply that would satisfy the purchase amount.
  3. The newly minted tokens go to the buyer, and the reserve currency stays locked in the contract's reserve pool.
  4. A seller sends tokens back to the contract; the contract burns them and calculates the reserve amount to return based on the same curve, always at or below the current price.
  5. Because every purchase moves the price up the curve and every sale moves it down, the price is fully determined by supply at all times — there's no external price discovery needed.
  6. Some designs (like Bancor's) keep only a fraction of raised funds as "reserve ratio" backing, using the ratio to control price sensitivity to trade size.
  7. Anyone can enter or exit at any time without waiting for a counterparty, since the contract is the counterparty.

Why designers use it

  • Guarantees instant, permissionless liquidity for a brand-new token with no need to bootstrap an order book or find market makers.
  • Makes price transparent and formula-driven rather than subject to manipulation by a single large trade against thin liquidity.
  • Lets early supporters buy in cheaply and rewards them for taking on early risk, since price rises monotonically with adoption.
  • Useful for continuous fundraising or "continuous organizations" where a project wants ongoing capital inflow tied directly to token demand.

Failure modes

  • Front-running and sandwich attacks: because price impact is fully predictable from the formula, bots can buy right before a large trade and sell right after, extracting value from ordinary users.
  • Death spirals: if the curve is steep and confidence drops, a wave of selling burns tokens and craters price rapidly, and there is no external buyer to slow the fall.
  • Reserve drain: aggressive early buyers can lock in cheap tokens, then dump on later buyers who pay far more, transferring wealth from latecomers to insiders.
  • Poor formula choice: an overly steep curve makes the token nearly untradeable at scale (huge slippage), while an overly flat curve fails to reward early risk-taking.
  • Reserve mismanagement: if the contract's reserve isn't properly isolated or is redirected (e.g., to a treasury) rather than fully backing redemptions, sellers may find they can't exit at the price the curve implies.

What to check before using it

  • Simulate the curve's behavior under both mass buying and mass selling scenarios, not just steady, moderate trading.
  • Confirm the reserve ratio and what percentage of funds are actually held to back redemptions versus diverted elsewhere.
  • Check for slippage and price-impact limits so a single large trade can't be used to manipulate downstream systems (e.g., using curve price as an oracle).
  • Verify whether the curve formula is upgradeable by an admin key, which would let a project change the rules retroactively.
  • Compare against a simple AMM pool: a bonding curve only makes sense when you specifically need mint/burn dynamics, not just liquidity.

Experiments that used it · 8

Shown oldest first, so you can watch the design evolve.

Bancor
The first on-chain automated market maker: a bonding-curve 'smart token' protocol that pioneered continuous, contract-native liquidity before Uniswap, later famous for its (ultimately paused) single-sided impermanent-loss protection.
2017 technically successful commercially unsuccessful
Curation Markets
Simon de la Rouviere's 2017 protocol-design framework for tokenized curation of information via continuous (bonding-curve) tokens — the concept never shipped as a single protocol but seeded bonding curves, TCRs, and continuous token models across the ecosystem.
2017 partial success
Fairmint (Continuous Organization)
Fairmint implemented Thibauld Favre's Continuous Organization model — a bonding-curve 'Decentralized Autonomous Trust' that continuously sells revenue-backed securities — before pivoting to legally simpler Rolling SAFEs and onchain cap-table infrastructure.
2019 technically successful commercially unsuccessful
Fei Protocol
A decentralized, ETH-backed stablecoin (FEI) that pioneered 'direct incentives' (mint rewards / burn penalties on DEX trades) and Protocol Controlled Value, but launched into a peg crisis and eventually wound itself down after a merger with Rari Capital and an $80M exploit.
2021 failed
BitClout
A 2021 blockchain social network that tokenized personal reputation via bonding-curve 'creator coins,' pre-loading thousands of celebrity profiles without consent, whose pseudonymous founder later revealed himself.
2021 technically successful commercially unsuccessful
Sudoswap
A minimal, fully on-chain automated market maker for NFTs that priced ERC721/ERC1155 tokens along customizable bonding curves and popularized royalty-optional, low-fee NFT trading.
2022 technically successful commercially unsuccessful
Friendzy
A Solana fork of friend.tech that tokenized X accounts into freely transferable 'Creator Keys' on a gentler bonding curve, did ~$1M volume in its first week, then bled to near-zero users and a lapsed domain within a year.
2023 abandoned
Meteora Dynamic Bonding Curve
Meteora's permissionless, configurable bonding-curve launch protocol on Solana that graduates tokens into a real AMM pool once a funding threshold is hit.
2025 ongoing