Onchain Atlas

Bonding Curve Launch

A way to launch a new token where the price rises automatically as more people buy, with no presale or fixed listing price.

Also called: bonding curve token launch · curve-based fair launch · pump-style launch

What it is

A bonding curve launch replaces a traditional token sale or exchange listing with a smart contract that mints and prices tokens algorithmically. Anyone can buy directly from the contract at a price determined by a formula (the "curve"), and the price increases as more tokens are bought and decreases as tokens are sold back. There's no order book and no need for a market maker to show up first — the curve is always the counterparty.

How it works

  1. A creator deploys a contract holding a mathematical curve (commonly linear or exponential) that maps total tokens issued so far to the current price.
  2. The first buyer pays the lowest price on the curve; the contract mints tokens to them and moves the cumulative supply counter forward.
  3. Each subsequent buy moves further along the curve, so price rises with every purchase — early buyers systematically pay less than later buyers.
  4. Funds paid in (often a base asset like ETH or a stablecoin) accumulate inside the contract as reserves, rather than going to the creator directly.
  5. Sellers can typically sell back into the curve at the current price, with the contract burning their tokens and refunding reserves according to the same formula, which keeps a floor price tied to the accumulated reserves.
  6. Many implementations set a graduation threshold: once the curve reaches a target market cap or reserve amount, the remaining liquidity is migrated to a standard AMM pool (like Uniswap) and the bonding-curve contract stops trading, "graduating" the token to normal exchange trading.

Why designers use it

  • Removes the need for a presale, VC round, or centralized listing — anyone can be the first buyer with no gatekeeper.
  • Guarantees continuous liquidity from the moment of launch, since the curve itself is always willing to buy or sell.
  • Aligns early support with lower entry price, rewarding people who believe in a project before it's proven.
  • Creates a transparent, deterministic price discovery process instead of an opaque initial listing price set by insiders.

Failure modes

  • If insiders or the creator buy first and dump into retail demand right after, later buyers absorb losses in a pattern sometimes called a "rug" even without any explicit fraud.
  • Bots that watch for new curve deployments can buy the first block of tokens within seconds, capturing most of the cheap supply before humans can act.
  • If the curve's reserve isn't fully backing outstanding supply (e.g., a portion of funds is skimmed as fees), the sell-back floor can be lower than buyers expect.
  • Thin, illiquid curves are highly volatile — a single moderately sized buy or sell can move price by large percentages, inviting manipulation.

What to check before using it

  • Read the exact curve formula and confirm whether reserves fully back the sellable floor price or whether fees are taken out first.
  • Check for anti-sniping protections (e.g., per-wallet caps, launch delays, or randomized start) if fair distribution matters.
  • Understand the graduation mechanics: what triggers migration to a standard AMM, and who controls that migrated liquidity.
  • Verify whether the creator or any insider wallet has a head start or special allocation before public trading opens.

Experiments that used it · 3

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

Luna by Virtuals
Virtuals Protocol's flagship AI-idol agent whose token launched via bonding-curve IAO on Base and who became the first AI agent to autonomously control her own socials, tip and hire humans onchain, and pay another AI agent in crypto.
2024 partial success
VaderAI
A Virtuals Protocol AI agent that tried to become the 'BlackRock of the Agentic Economy' via AI-managed investment DAOs with fee-funded token buybacks, then pivoted to embodied-data collection for physical AI after its token collapsed ~98% from its January 2025 peak.
2024 technically successful commercially unsuccessful
Believe (Launchcoin)
Solana launchpad that let anyone create a token by replying to an X account, branded as 'Internet Capital Markets' for startups — it minted thousands of coins a day at its peak before activity fell away sharply.
2025 failed