.jpg?table=block&id=3982d44d-39b0-8067-acb1-d118e71a6b99&cache=v2)
Table of Contents
- Introduction
- Not every chain needs the same stack
- 1. Stablecoins: the first integration, not the tenth
- 2. Oracles: lending cannot launch without them
- 3. The DEX: build, fork, deploy, or license
- 4. Automated liquidity management: what keeps pools alive
- 5. Lending: permissionless rails first, flagship deployments later
- 6. Perps: rent depth, don't bootstrap it
- 7. Interop: two rails minimum
- 8. Data, indexing, explorers: distribution is technical
- 9. Wallets and accounts: free to support, negotiated to be default
- 10. Onramps, custody, and the institutional gate
- 11. Liquid staking, for L1s
- 12. Risk, security, and the operational layer
- What actually launched: the receipts
- Sequencing: what is permissionless and what is a negotiation
- The part pre-deposits cannot buy
- About Algebra
Do not index
Introduction
Forty-plus mainnets launched between early 2025 and mid 2026 - from Tempo and Arc to Robinhood Chain and Monad. Blockspace is a commodity now. What separates chains that retain users from ghost towns is the financial infrastructure that is live, liquid, and usable on day one. This is the map: every category, the credible providers, and how launches actually get them.
The profile of who launches a blockchain changed in the last eighteen months. Robinhood shipped an L2 for tokenized stocks with Uniswap, Chainlink, and BitGo live on day one. Stripe and Paradigm incubated Tempo, which reached mainnet in March 2026 with Visa and Stripe running validators and Deutsche Bank, Revolut, and OpenAI as design partners. Circle's Arc opened its testnet to more than a hundred institutions, including BlackRock, Goldman Sachs, HSBC, Visa, and Mastercard. Google is building GCUL with a CME pilot. On the crypto-native side, Monad went live with 300+ projects and a $188M Coinbase public sale, and Plasma launched with $2 billion in stablecoins already on the chain.
Execution layers are increasingly interchangeable. Berachain launched in February 2025 with $3.1 billion in pre-deposited liquidity and lost roughly 88% of its TVL within a year. Pre-deposits buy a launch headline. Infrastructure is what retains.
This guide is written for the people making these decisions: chain foundations, BD leads, and protocol teams deciding where to deploy next.
Not every chain needs the same stack
The biggest mistake in ecosystem planning is copying another chain's checklist without asking who the chain is for. The 2025-26 cohort splits into four archetypes, and their day-one priorities are almost inverted.
Stablecoin / payments
Tempo · Arc · Plasma · Stable
First: issuance, fiat ramps, gas UX,
compliance rails
Later: perps, LSTs, long-tail DeFi
Trading / consumer fintech
Robinhood Chain · Katana
First: deep spot liquidity, low-latency oracles, perps, market-maker access
Later: broad app diversity
Institutional / RWA
Arc · Converge · GCUL
First: qualified custody, permissioned lending, KYC assets, RWA oracles
Later: public perps, farming
General-purpose DeFi
Monad · MegaETH
First: DEX, lending, oracles, bridges, stablecoin liquidity
Later: custody, compliance tooling
Four chain archetypes of the 2025-26 launch cohort - day-one priorities are almost inverted between them.
The institutional column deserves emphasis because it inverts the usual order. A chain whose users are banks and asset managers cannot onboard a single customer until a qualified custodian supports it. Fireblocks or Anchorage support matters more to Arc than any DEX does. Even Monad, the most crypto-native launch of the cohort, had Anchorage custody confirmed before mainnet and ran its public sale through Coinbase. The direction of travel is one way: everything is moving closer to institutions.
The chain launch stack, bottom up
STABLECOINS & GAS UX · Bridged USDC Standard, USDT0, AUSD/M0, native issuance
ORACLES · RedStone, Chainlink Scale, Pyth, Chronicle
SPOT DEX + ALM · canonical Uniswap / v3 fork / custom / licensed engine (Algebra)
LENDING · Morpho stack, Aave (+ Horizon), Euler EVK · then PERPS via Orderly / GMX
INTEROP · LayerZero, Hyperlane, Wormhole NTT, CCIP + intents bridges
DATA · Blockscout explorer, Goldsky/Envio indexing, Dune · WALLETS · Safe, WalletConnect, Privy
ACCESS · fiat ramps (MoonPay, Transak) · custody (Fireblocks, Anchorage, BitGo) · risk (Chaos Labs)
Each layer depends on the one below it. Lending cannot launch without oracles; oracles are useless without dollars to price.
1. Stablecoins: the first integration, not the tenth
Nothing else works without dollars. Lending needs collateral and a borrow asset. DEX pools need a quote asset. Users need something to hold that is not the (volatile, probably unlisted) gas token. The menu in 2026 is well developed:
- Bridged USDC Standard. Circle's specification lets a new EVM chain deploy canonical bridged USDC at launch, with a pre-wired path for Circle to later take over the contract and flip it to native issuance with no liquidity migration. This is the self-serve default for new L2s; World Chain launched with it and upgraded later. Native USDC plus CCTP V2 (30-second cross-chain transfers) comes afterward, as a negotiated deal.
- USDT0. Omnichain USDT built on LayerZero's OFT standard: USDT locked on Ethereum, USDT0 minted on the destination chain. It moved over $63 billion in its first year, and it is how Plasma launched liquid in September 2025, with stablecoins active from block one across more than a hundred DeFi partners.
- White-label dollars. Agora's AUSD and M0's extension model let a chain launch its own branded or compliance-configured dollar on shared liquidity rails. Ethena runs the same play at larger scale: MegaETH's native USDm is issued via Ethena on USDtb rails, with reserves largely in BlackRock's BUIDL fund, and the float income subsidizes the chain's sequencer. The stablecoin is the business model, not an accessory.
The payment chains go further and enshrine this layer. Tempo built a native Fee AMM so gas is payable in any stablecoin, plus a TIP-20 token standard with built-in transfer policies. Arc uses USDC as gas natively. For institutional chains, the stablecoin is not an integration; it is the chain's reason to exist.
Launch pitfall: gas UX is stablecoin infrastructure
Stable launched in December 2025 with $1.1 billion pre-deposited - then users struggled to move funds because there was no simple way to acquire the gas token. The team shipped conversion tools after the fact. If your chain uses a custom gas asset, the swap path into it must exist before users arrive, not after the complaints do.
2. Oracles: lending cannot launch without them
Price feeds gate everything credit-shaped. No feed, no money market, no perps, no liquidations. Two delivery models matter: push oracles write prices on-chain on a schedule and are what lending protocols consume; pull oracles sign prices off-chain and let anyone post the update in the transaction that uses it - cheaper, and suited to derivatives. A serious chain wants both.
Provider | Model | What the record shows |
RedStone | Push + pull, modular | |
Chainlink | Push + streams + CCIP | Incumbent standard; Scale program = foundation subsidizes costs for day-one feeds (Plasma, Soneium); Robinhood Chain's chosen oracle; compliance / proof-of-reserve stack is the institutional draw |
Pyth | Pull, permissionless | Deploy the receiver contract, get 1,300+ first-party feeds; Lazer targets ~1ms for perps venues |
Chronicle | Push, gas-optimized | Ex-MakerDAO team; cheapest updates; verification for BUIDL, Superstate, Centrifuge - natural fit for RWA chains |
Sequencing note: oracle deployment now takes days to weeks and routinely lands before mainnet. Secure it before you announce lending partners, because they will ask.
3. The DEX: build, fork, deploy, or license
Spot liquidity is the center of gravity for any chain that wants organic activity. It is where new tokens get markets, where aggregators route, and where the chain's TVL either works or sits idle. A chain team has four real options in 2026, and the honest answer is that most serious chains use more than one.
Path | What you get | What it costs | The fine print |
Canonical Uniswap | Real distribution: web app, wallet, API route in day one (Monad, MegaETH, Robinhood Chain) | DAO process, not a sales call; no dedicated frontend/BD from Labs by default | v3 via DAO + Accountability Committee; v4 under BUSL until June 15, 2027 - production needs a governance license grant |
Fork Uniswap v3 | Battle-tested code, free since the license went GPL in April 2023; every ALM/aggregator understands it | You own the whole build: contracts, frontend, indexer, farming, fresh audits on changes | 2021 feature set - static fee tiers, no hooks, no native dynamic fees |
Custom AMM | Maximum differentiation | Longest timeline, fullest audit burden | Justified when the AMM design is the product (Curve's stableswap; Tempo's enshrined Fee AMM) |
Licensed engine (Algebra) | Algebra Integral: audited modular CLAMM core + stock plugins (dynamic fees, farming, limit orders, oracles); DEX owns brand, fees, token | Commercial license; a vendor relationship instead of a free fork | Algebra's figures: 100+ DEX launches, 50+ chains, $200B+ cumulative volume (DeFiLlama fork data); Camelot, QuickSwap, THENA, Lynex built on it |
Two factual distinctions between the licensed-engine model and deploying v4. First, Algebra's plugins can be added or upgraded on live pools without migrating liquidity, whereas a Uniswap v4 pool's hook is fixed at pool creation. Second, ownership: a licensee keeps its brand, fee take, and token economy, where a canonical deployment routes value through the Uniswap ecosystem.
The strategic point: canonical blue-chip deployments and a native venue are complements, not substitutes. Uniswap's presence gives traders a familiar venue and aggregators a route. A native DEX - custom-built or licensed - is what keeps fee flow, incentives, and governance inside the ecosystem. Katana went furthest, curating a single spot venue (Sushi) and recycling chain fees back to its apps. Most chains land in between.
Whatever the venue, aggregator integrations (1inch, 0x, ODOS, ParaSwap) belong on the launch checklist, not the roadmap. A DEX that aggregators cannot route into is invisible to a large share of flow.
4. Automated liquidity management: what keeps pools alive
Concentrated liquidity is capital-efficient and operationally demanding. LPs pick ranges, drift out of them, stop earning, and leave. On a new chain with volatile early markets this happens fast.
ALM protocols - Gamma, Steer, ICHI, and Krystal are the established names, with coverage across Uniswap-style and Algebra-based deployments - wrap positions in managed vaults that rebalance automatically. The practical consequence for a chain team: your liquidity mining budget works several times harder when incentives point at managed vaults instead of raw positions, because deposits stay in range and passive LPs can actually participate. Line up at least one ALM before incentives go live.
5. Lending: permissionless rails first, flagship deployments later
Lending is the second-largest DeFi vertical and the clearest illustration of the split between negotiated and permissionless infrastructure.
Aave is governance-gated. A new-chain deployment needs a forum proposal, risk-provider signoff, a vote, and usually incentives from the host chain: Sonic's deployment came with a $15M+ package. Plasma's launch collaboration shows what a priority deal looks like: Aave live early, $6.2 billion deposited within about a week. For institutional chains, Aave Horizon is the permissioned instance where qualified institutions borrow stablecoins against tokenized funds - the reference design for KYC-gated credit, and a named core app on Converge.
Morpho is the fast path. Its "Morpho Everywhere" model deploys the immutable lending stack to new EVM chains as pure infrastructure, with local teams building curated vaults and frontends on top. That is how a dozen chains got credible lending in 2025 without any governance negotiation, and it is behind more distribution than most people realize: Coinbase's USDC lending, World App, Ledger and Trust Wallet earn products, the Morpho-powered vaults inside the Robinhood app, and Katana's launch stack. Plume's native money market, Mystic, is a branded curator UI over Morpho markets - exactly how an RWA chain gets compliant credit without forking anything. Morpho V2's fixed-rate, fixed-term loans target the institutional cohort directly.
Euler v2 deserves a mention for the same reason: its vault kit is permissionless and supports custom or permissioned vault configurations, useful when neither an Aave campaign nor a Morpho curator ecosystem exists yet.
The pattern: deploy permissionless rails at genesis, recruit one or two serious vault curators (this matters more than the deploy itself), and treat a flagship Aave deployment as a milestone to negotiate once there is usage to point at.
6. Perps: rent depth, don't bootstrap it
Onchain perps did roughly $12 trillion in volume in 2025, about triple the prior year, and almost none of the winners bootstrap orderbooks chain by chain anymore. A new chain has three realistic routes:
- Shared-liquidity infrastructure. Orderly Network runs an omnichain central limit orderbook on LayerZero rails; a chain deploys settlement contracts and any local team can launch a branded perps venue with shared depth from day one. Vertex Edge runs a synchronous orderbook across EVM chains with settlement on the user's origin chain. This is the fastest credible path.
- Per-chain deployments of pool-based venues. GMX expanded to Solana and the Bitcoin L2 Botanix in 2025 through DAO-governed proposals. Chains court these deployments the way they court Aave.
- A dedicated integration. Robinhood Chain routes perps through Lighter via the Robinhood Wallet. Converge's venue is Ethereal, aligned with Ethena. This is a BD outcome, not an off-the-shelf product.
Two cautionary notes. dYdX, the previous generation's answer, lost most of its share to Hyperliquid and is not deploying to your chain. And Elixir, which supplied orderbook liquidity to several venues, shut down its deUSD synthetic dollar in November 2025 after the Stream Finance collapse. Diligence your liquidity partners like counterparties, because that is what they are.
Worth knowing even though it is the opposite of a chain integration: Hyperliquid's HIP-3 lets teams stake HYPE to launch their own perp markets on Hyperliquid itself. Some projects that would have launched an appchain now launch a market instead. That is your competition for derivatives flow.
7. Interop: two rails minimum
- LayerZero is the gravity well, connecting 165+ chains, because the assets a new chain wants arrive as OFTs: USDT0, AUSD, USDe. If your stablecoin plan involves any of those, LayerZero connectivity is implied.
- Hyperlane is the only major stack a team can self-deploy without anyone's approval - the default for appchains and rollups needing connectivity at genesis. The cost: you run your own relayer.
- Wormhole brings the strongest non-EVM coverage and its Native Token Transfers standard; Monad's native bridge runs on it.
- Chainlink CCIP typically arrives bundled in Scale agreements and carries weight with institutions standardizing on Chainlink.
The observed 2025-26 pattern for a general-purpose chain: LayerZero plus Hyperlane at launch, CCIP if in Scale, an intents-based fast bridge (Across, Relay) for UX on top of the canonical bridge, and Wormhole where Solana matters. Monad launched with four interop providers simultaneously. That is not excess; bridges are how users arrive, and each rail carries different assets.
8. Data, indexing, explorers: distribution is technical
Integrators do not read your docs out of curiosity. If pools are not indexed, wallets do not show positions, aggregators do not route, and grant committees cannot measure anything.
The stack has commoditized in a useful way. Blockscout's Autoscout gives any EVM chain a launch-day explorer with Etherscan-compatible APIs, with a paid Etherscan instance as a later brand upgrade. Goldsky and Envio index custom chains without any approval process, which made them the day-one choice for most of the 2025 cohort; The Graph's decentralized network involves a governance integration process that arrives later. A Dune integration (100 chains supported as of early 2025) is table stakes for public analytics and the reporting your grant program will demand.
None of this is glamorous, and all of it is why Alpen, a Bitcoin L2 that has not even reached mainnet, already lists Alchemy, Tenderly, and SubSquid on its ecosystem page alongside Morpho, RedStone, LayerZero, and Anchorage. Pre-announcing the boring infrastructure is what serious launch preparation looks like.
9. Wallets and accounts: free to support, negotiated to be default
MetaMask and Rabby support any EVM chain through custom RPC entries, so baseline support costs nothing beyond registry listings (chainlist, EIP-3085 metadata). Default listing - the difference between "users can add your chain" and "users see your chain" - is curated and negotiated. Budget BD time for it.
Three others belong on the checklist. WalletConnect onboarding, since it is the session layer for most of the ecosystem. An embedded-wallet provider (Privy, now Stripe-owned, or Dynamic) if the chain targets consumer or payments flows where seed phrases are a non-starter. And Safe, which is not optional: protocol treasuries, market makers, and every serious team refuse to operate from EOAs. Safe contracts at genesis, with a hosted interface, is a quiet prerequisite for the entire ecosystem above it.
10. Onramps, custody, and the institutional gate
Fiat ramps (MoonPay, Transak, Ramp) integrate at the chain level: one integration covering the gas token and lead stablecoin unlocks embedded ramps for every app on the chain. Payments-focused chains should weight offramp coverage and local payment methods more heavily than listings.
Custody is the institutional gate, and it is binary. Market makers, funds, and corporate treasuries cannot hold assets on your chain until Fireblocks, Anchorage, BitGo, or Copper supports it. Fireblocks support is the single biggest unlock for market-maker depth. Anchorage now announces support before mainnets to capture launches; it did so for Monad's MON and was day-one custodian for Stable. Robinhood Chain launched with BitGo. Botanix launched with Fireblocks in its federation. For RWA and bank chains this is not a checklist item, it is the checklist: Arc's testnet cohort of banks and asset managers implies custody at genesis or no users at all.
11. Liquid staking, for L1s
A proof-of-stake L1 wants an LST live roughly when lending launches, because staked-asset collateral is what makes the DeFi flywheel turn. The realistic 2026 approach is one or two native LST providers recruited pre-launch. The counterexample is instructive: Lido discontinued official wstETH support on nine chains in June 2026 after usage failed to materialize. Logo integrations that nobody uses get pruned. Institutional chains often skip or permission this layer entirely, for regulatory reasons as much as demand.
12. Risk, security, and the operational layer
The last category is the one nobody tweets about. Chains that court institutions now retain risk firms the way protocols retain auditors: Chaos Labs appears on Alpen's ecosystem page under its own category, and Aave will not enter a chain without risk-provider signoff. Beyond that: audits on everything users touch, a bug bounty at launch, oracle and TWAP monitoring, admin-key transparency, an incident-response plan that names humans, and a liquidity migration plan you hope never to use.
The teams that do this well behave less like app launchers and more like market operators. The teams that do not become case studies.
What actually launched: the receipts
Theory aside, this is what day one looked like across the recent cohort:
Chain (launch) | Stablecoins | DEX | Lending | Oracle | Interop | Custody |
Tempo (Mar 2026) | Multi-issuer, stablecoin gas | Enshrined Fee AMM + native DEX | - | - | - | Zodia (validator) |
Robinhood Chain (Jul 2026) | USDG | Uniswap primary | Morpho vaults, Earn | Chainlink | Arbitrum stack | BitGo |
Arc (testnet) | USDC native + gas, EURC, USYC | Native FX/swap infra | - | Chainlink* | Circle CCTP/Gateway | 100+ institutions onboard |
Monad (Nov 2025) | Native + bridged | Uniswap day one | Morpho, ecosystem | Chainlink, Pyth, RedStone | Wormhole, LayerZero, Axelar, Hyperlane | Anchorage |
Plasma (Sep 2025) | USDT0, $2B day one | Ecosystem venues | Aave, Fluid, Euler | Chainlink (Scale) | LayerZero | - |
MegaETH (Feb 2026) | USDm (Ethena rails) | Uniswap + GTE | Avon, ecosystem | RedStone Bolt | - | - |
Stable (Dec 2025) | USDT0, gUSDT gas | - | - | - | LayerZero | Anchorage |
Katana (Jun 2025) | Curated | Sushi | Morpho | - | AggLayer, VaultBridge | - |
Alpen (pre-mainnet) | BTD (Liquity), Circle, Agora | - | Morpho, Fluid, Tenor | RedStone | LayerZero, Garden | Anchorage |
- Arc developer tooling per secondary sources; Circle's PR lists categories without full vendor enumeration.
Patterns worth stating plainly. Morpho and Uniswap show up almost everywhere DeFi does. Chainlink and RedStone split the oracle market along institutional and speed lines. LayerZero connectivity is close to universal because the stablecoins ride on it. Anchorage turned pre-mainnet custody announcements into a repeatable playbook. And the payment chains (Tempo, Arc) enshrine or internalize what DeFi chains outsource.
Sequencing: what is permissionless and what is a negotiation
The single most useful planning distinction:
Deployable in days — no permission needed
- Hyperlane interop (self-deploy)
- Pyth receiver contracts
- Morpho lending stack
- Euler vault kit
- Safe smart accounts
- Blockscout explorer (Autoscout)
- Goldsky / Envio indexing
- Bridged USDC Standard
- Baseline wallet support (public registries)
Negotiated — weeks to months, often with incentives
- Chainlink Scale · native USDC · USDT0 at size
- Aave deployment (governance + incentives)
- GMX / perps venue deployments
- Canonical Uniswap (v4 grant before mid-2027)
- Licensed DEX engine (Algebra)
- Custody: Fireblocks, Anchorage, BitGo
- Dune · The Graph · default wallet listings
- Market-maker commitments · fiat ramps
Chains that launch well ship the entire permissionless set at genesis and time the negotiated set so the flagship announcements land when users are actually there to use them.
The part pre-deposits cannot buy
Plasma set the "launch liquid" bar. Berachain proved that billions in pre-committed TVL can evaporate when the incentives that attracted it decay. Stable showed that a billion dollars of deposits does not help if users cannot pay for gas. The common thread: liquidity events are marketing; infrastructure is retention.
The chains holding attention a year after launch are the ones where a user can arrive, fund a wallet in one step, swap at reasonable depth, put assets to work in credit markets, verify all of it on an explorer, and leave without friction - and where an institution can do the same through a custodian with a compliance officer's blessing. Every piece of that experience is a named vendor, a contract deployment, or a negotiation with a lead time measured in months.
Launching the chain is the easy part now. The ecosystem is the product.
Sources are linked inline throughout. Figures attributed to protocol teams (e.g. Algebra's 100+ DEX launches / $200B+ volume, Plasma's day-one liquidity) are first-party and cross-referenced where independent data exists (DeFiLlama, Dune). Compiled July 2026.
About Algebra
Algebra Labs develops advanced Automated Market Maker infrastructure for decentralized exchanges, enabling higher capital efficiency, flexible monetization, modular upgradability, and faster product iteration via CLAMM & modular architecture — V4 plugins + light core.
Algebra powers 100+ DEXs across EVM chains, including Blackhole, Camelot, THENA, Supernova, QuickSwap, Hydrex, SparkDEX, SwapX, StellaSwap, Nest Exchange, and more.
Learn more on our website: algebra.finance.
