Launching a meme

Leave all three alone and you get this deployment's default launch, which is what most memes want. The form fills them in for you and the panel on the right shows what you are about to sign.

A bonding curve is a contract that is always willing to trade. There is no order book and no counterparty: the price comes from a formula, it rises as tokens leave the contract and falls as they come back. Everyone buys from the same line, in the order their transaction lands, including the creator.

The rule is a constant product. Call the starting market cap V₀ and the ADA bought in so far R: the contract keeps (V₀ + R) × (tokens still in the curve) constant. Three consequences follow, and the rest of this page is just those three:

• Market cap is (V₀ + R)² ÷ V₀ — so the cap climbs much faster than the money going in.
• The share of the supply sold is 1 − √(V₀ ÷ cap) — the curve approaches the whole supply but never reaches it.
• Reaching a cap M takes √(M × V₀) − V₀ of buying. The supply is not in that formula.

The starting market cap is the curve's virtual reserve: ADA the contract pretends it already holds. Nobody pays it, it is not a fee, and it never lands in the pool. It exists because a constant-product curve with nothing on the ADA side has no first price — the first buyer would take the entire supply for approximately nothing. The virtual reserve is what gives token number one a price.

• It sets the opening price: starting cap ÷ supply. At 2,000 ADA and a billion units that is 0.000002 ADA per token.
• It sets the climb, because the climb is graduation cap ÷ starting cap, and the climb is what buyers came for.
• It sets the burn, which is the same ratio upside down. Section 5.
• A higher start is a calmer launch — less multiple, less burn, more ADA needed to graduate. A lower start is the opposite.

This deployment allows 1,000 to 3,000 ADA, and the contract refuses anything outside that range — the website does not get a say.

The graduation cap is where the curve closes and the liquidity locks. You can type it as a market cap or as the share of the supply the curve should sell; they are the same dial, because selling a share f is exactly a climb of 1 ÷ (1 − f)².

Read the last two columns together: a 2,000 ADA token that graduates at 100,000 ADA — a fifty-fold climb — needs 12,142 ADA of actual buying, not 100,000. The gap is the virtual reserve doing its work, and it is why a market cap is not money anybody holds.

Supply is missing from that formula on purpose. Minting a million units or a hundred billion changes the price per token and nothing else: the same 12,142 ADA graduates either one.

When the last ADA needed arrives, the curve closes and one transaction moves everything into an AMM pool with no owner, no LP tokens and no withdrawal path. It puts in two things:

Worked through: if 5,000 ADA was raised and the final price is 0.01 ADA, the pool opens with 5,000 ADA and 500,000 tokens — because 5,000 ÷ 0.01 = 500,000. The first price in the pool is then identical to the last price on the curve. Pair less and the token would open above the curve; pair more and it would open below, which is a gap a bot takes within one block.

A curve never sells its whole supply — it approaches it. At graduation there are still tokens inside, and only some of them can go into the pool (section 4). The remainder is destroyed by the same transaction, and the share is exact:

burned = starting cap ÷ graduation cap = 1 ÷ climb2,000 ÷ 100,000 = 2.0% of the mint, which is one divided by 50×.

There is no third option for those tokens, and each alternative fails for its own reason:

• Leave them in the contract — they would be an overhang above every holder, waiting for the next rule change.
• Give them to the creator — that is a team allocation, which is the thing this platform exists to not have.
• Put them all in the pool — the pool price would open below the curve's last price, and the launch would end with a drop.

The practical reading: a bigger climb burns less. At 5× roughly 20% is burned; at 100× it is 1.0%.

The same launch can be described by three different numbers, and the form lets you type whichever one you actually care about; it derives the mint from it. Below: you type 1,000,000,000 in each mode, on the 2,000 → 100,000 ADA curve above.

• sold on the curve — the number buyers take off the curve.
• that exist after graduation — what is left once the burn has happened.
• minted at launch — the number the minting policy creates, once, forever.

So "I want ten million tokens afterwards" means minting more than ten million and burning the difference. It is not waste: the burned tokens never belonged to anybody, and the number you asked for is the number that survives. The percentages in the table come straight from the two caps — nothing here is a choice the platform makes.

The figures above ignore the graduation fee, which comes off the ADA side and therefore nudges the burn up by a fraction of a percent. The simulator in the form shows the real split, fee included.

The mint happens once and can never happen again: the minting policy is parameterised by one UTxO of your wallet, and the launch transaction spends it. The detail is here →

The first minutes of a launch are when a bot can take the whole curve in one transaction. Two caps stand in the way, and yours can only make the launch stricter than the platform's, never looser.

A cap per buy is really a statement about how many buyers you want. With 12,142 ADA needed, a cap of 100 ADA means at least 122 separate buys to fill the curve; a cap of 1,000 ADA means at least 13. The form lets you type it either way round, and shows both.

Note what the window does not do: it caps the size of one buy, not the number of wallets anyone controls. It raises the cost of taking a launch; it does not make it impossible.

A sale can be announced for a moment in the future, up to 7 days ahead. The token mints and locks immediately, waits on the Upcoming page with a countdown, and the contract refuses every buy until the moment arrives — yours included. Your fair-launch window then starts counting from the opening, not from the mint.

Read live from the protocol's settings UTxO. The bounds are checked by the validators, not by this website, and the values your token launches with are frozen in its own datum — a later change to these settings never reaches a token that is already running.