01 · BasicsWhat a Dynamic AMM is
A Dynamic AMM is Meteora’s take on the classic constant-product pool. Two tokens sit in reserves, price follows x·y = k, and every LP shares in every trade. The “dynamic” part is everything layered on top: how fees behave, where fees are paid, and what happens to the liquidity itself.
If you have used any AMM, the core is familiar. When a trader buys token A with token B, the pool’s A reserve shrinks, its B reserve grows, and the price of A rises along the curve. Full-range liquidity never goes out of range, so a position keeps earning whatever the price does. That makes DAMM pools the most passive option on Meteora and a natural fit for brand-new tokens whose fair price nobody knows yet.
Meteora has shipped two generations. DAMM v1 is now listed under legacy products. DAMM v2, internally called cp-amm, is a brand-new program, not an upgrade of v1, and is the version Meteora recommends for new pools. At the time of writing DAMM v2 hosts a very large number of pools, many of them small launch pools, which tells you where its center of gravity is: token launches and the long tail of new assets.
02 · VersionsDAMM v1 vs DAMM v2
The two versions share a name and a curve family but solve different problems. Version one was built around yield on idle reserves; version two is built around launches and fee control.
| Feature | DAMM v1 (legacy) | DAMM v2 |
|---|---|---|
| Position format | LP token | Token-2022 NFT |
| Lending yield on reserves | Yes | No |
| Stable-swap curve | Yes | No |
| Concentrated price range | No | Optional |
| Fee scheduler / anti-sniper | No | Yes |
| Fees in quote token only | No | Yes |
| Locked liquidity with fees | Yes | Yes |
| Built-in liquidity mining | External farms | Up to 2 tokens |
Fees tell the story too. A standard DAMM v1 constant-product pool charges 0.25%, split 20% to the protocol and 80% to LPs, and v1 stable pools charge 0.01%, all to LPs. DAMM v2 has no single fee: each pool’s config sets it, anywhere from 1 bps up to a 99% cap on current pools (older v0 pools are capped at 50%). The protocol takes 20% of the trading fee by default, and when a referral account is passed, the referrer receives 20% of that protocol share.
03 · Anti-snipingFee schedulers and anti-sniper fees
The first seconds of a token launch belong to bots. A fee scheduler makes those seconds expensive: the swap fee starts very high and decays, so a sniper who buys in the first block pays a heavy toll that goes to LPs.
DAMM v2 offers four base-fee scheduler modes for new pools:
Time scheduler
The fee falls step by step as time passes since activation. In linear mode it drops by a fixed amount each period; in exponential mode it drops by a percentage-style factor each period, so it falls fast at first and then flattens.
Market-cap scheduler
The fee falls as the token’s price rises above its starting level, again linearly or exponentially. It stays high until price passes the configured steps or the scheduler expires, which ties the discount to real demand rather than the clock.
An older rate-limiter mode, which raised fees on larger buys, is deprecated for new configs; existing pools using it can still trade. On top of any base mode, a pool can add an optional dynamic fee that reacts to volatility, much like the variable fee in DLMM. The total is always min(base + dynamic, pool cap).
Reading the widget
The fee scheduler widget at the top of this page uses a 50% starting fee, a 1% ending fee and a 60-minute window by default. With the exponential curve, the fee is already far below its starting level by minute 15, and the average across the window is much lower than the headline 50%. Switch to linear and the early fee stays punishing for longer. Real pools are configured in discrete periods rather than a smooth line, and every launch chooses its own numbers, so treat the widget as an intuition tool rather than a quote.
04 · CollectionQuote-only fees and compounding
Where fees land matters as much as how big they are. DAMM v2 supports three collect-fee modes:
- BothToken: fees accrue in whichever token was paid in, as in a classic AMM.
- OnlyB: all fees are collected in token B, usually the quote asset such as SOL or USDC. Most launch and memecoin pools use this mode.
- Compounding: a full-range constant-product mode in which
compounding_fee_bps(0 to 10,000) sets how much of the LP fee is reinvested into the reserves. The rest stays claimable in token B; a value of 0 leaves the whole LP fee claimable.
Quote-only collection is a quiet but important feature. An LP or launch team in a new token does not want to be paid in that token, because it may be worth a fraction tomorrow. Being paid in SOL or USDC turns trading activity into hard income that does not depend on the token holding its value. It is also why projects can permanently lock their liquidity and still fund themselves from trading fees.
DAMM v2 also supports an optional concentrated mode: the pool creator sets a fixed minimum and maximum square-root price, and liquidity is active only inside that band. This raises capital efficiency, but exact-in swaps that would push price past the boundary fail unless the partial-fill path is used. Meteora’s own MET token launched with a single-sided DAMM v2 pool spanning $0.50 to $7.50, with a fee scheduler that started high and declined, a clear example of a concentrated launch range.
05 · OwnershipPosition NFTs and permanent locks
Every DAMM v2 position is a Token-2022 NFT. Holding the NFT means controlling the position, and transferring the NFT transfers the liquidity with it. Positions can also be split, which lets an owner carve a tranche off for a partner or sell part of a position without touching the rest.
| Unlocked | Withdraw any time |
|---|---|
| Vesting | Unlocks gradually on a schedule |
| Permanently locked | Can never be withdrawn |
| Fees on locked liquidity | Still earned and claimable |
| Rewards | Up to 2 tokens, pro rata to liquidity |
| Token support | SPL and Token-2022 |
Permanent locks are the answer to the oldest launch scam in DeFi: the team adds liquidity, buyers pile in, and the team withdraws the liquidity, leaving holders with a token nobody can sell. If the launch liquidity is permanently locked in DAMM v2, that specific move is impossible, and the team can still earn fees from it. Launchpads built on Meteora use this to give buyers verifiable assurance.
06 · LegacyDynamic Vaults and the v1 yield layer
The original pitch of DAMM v1 was that idle liquidity should not be idle. Each side of a v1 pool is a position in a Dynamic Vault: a single-asset vault in which an off-chain keeper called Hermes allocates funds across approved lending strategies, up to 29 per vault. LPs earned swap fees plus lending yield on the part of the reserves not needed for trading. Vault profit unlocks gradually, and the performance fee is minted as vault LP tokens.
That linkage is what made v1 “dynamic.” Today, Meteora lists Dynamic Vaults among its legacy products. The docs note that most older strategy types now return “unsupported,” with Jupiter Lend the only external strategy currently implemented for active deposits. Vault TVL is still meaningful, so existing users are not stranded, but new liquidity design has moved on.
DAMM v2 deliberately drops vault yield. Keeping the full reserves inside the pool makes the accounting simpler and removes a layer of lending-protocol risk, while the new fee tools give LPs other ways to earn. DLMM has never been linked to Dynamic Vaults either.
07 · DecisionWhen to choose DAMM v2 and when DLMM
Both are core Meteora products, and they are complements rather than rivals. A simple decision frame:
Pick DAMM v2 when
- You want a passive, full-range position
- You are launching a token and need locked liquidity
- You want fees paid in SOL or USDC
- You need anti-sniper fee schedules
- You want a transferable NFT position
Pick DLMM when
- You want to concentrate capital around price
- You want Spot, Curve or Bid-Ask shapes
- You want one-sided, order-like ranges
- You want fees that rise with volatility
- You can monitor and rebalance
For an LP, the honest comparison is effort against efficiency. A full-range DAMM v2 position follows the textbook impermanent-loss curve and needs almost no maintenance. A DLMM position can earn far more per dollar while price stays in range, but it can also sit idle once price leaves. Many experienced LPs hold both: a DAMM v2 core in pairs they want to own for a long time, and DLMM ranges where they have a view on price.
Whichever you choose, the usual caveats apply. DAMM v2 has been audited by OtterSec, Offside Labs and Zenith, and its programs can be upgraded by Meteora. Audits reduce smart-contract risk but cannot remove it, and no fee mode protects you from a token that loses its value.
