What is Chainlink? A beginner's guide to LINK

Chainlink is the underappreciated infrastructure layer underneath most of DeFi. Almost every major lending protocol, derivatives platform, and algorithmic stablecoin relies on Chainlink to bring real-world prices on chain. Aave, Compound, MakerDAO, Synthetix, and dozens of others would simply not function without an oracle network providing reliable external data.

The story of why is the story of the oracle problem — one of the most important architectural problems in crypto, and one that almost no user ever thinks about directly. Yet failures of oracles have caused some of the largest single-day losses in DeFi history. Understanding oracles is understanding why DeFi works when it works, and why it breaks when it breaks.

The oracle problem in plain terms

A blockchain is, by design, a self-contained system. Smart contracts on Ethereum can see other transactions on Ethereum. They can read the state of other smart contracts. They cannot, on their own, see anything that exists outside the blockchain.

This is a feature, not a bug. The integrity of the blockchain depends on its ability to reach consensus about its internal state without trusting any external party. If a smart contract could just query a website for data, the security of any decision it made based on that data would depend on the website — and the consensus mechanism would lose its meaning.

But almost every useful decentralized application needs to know something about the outside world. A lending protocol needs to know the current dollar value of the collateral. A derivatives platform needs to know the price of the underlying asset. A weather-insurance smart contract needs to know whether it actually rained. A prediction market needs to know who won the election.

How do you get this external information into a smart contract without compromising the trust model of the blockchain itself? That is the oracle problem.

Why naive solutions don't work

The simplest possible oracle is: pick a trusted party, have them publish data to a smart contract, have other contracts read from there. This is centralized and breaks immediately under any adversarial pressure. The trusted party can be bribed, hacked, or simply make a mistake, and every decision downstream of that data is corrupted.

A slightly better version is to aggregate multiple sources and take the median or average. This is more resistant to single-source failures but introduces new problems. Who picks the sources? Who decides when a source has gone offline? Who responds when sources disagree?

The fundamental challenge is that you need a system that is decentralized in the same way the blockchain itself is decentralized, with no single point of failure, no party that can corrupt the data without consequence, and economic incentives that align participants toward providing accurate data.

This is exactly what Chainlink built.

How Chainlink works

Chainlink is a decentralized network of independent node operators, each of which is responsible for fetching real-world data and publishing it to smart contracts on supported blockchains. The architecture has several layers:

Node operators are independent companies and individuals who run Chainlink nodes. They source data from various external APIs, normalize it, and publish to the network. There are several hundred independent operators globally, including Deutsche Telekom, Swisscom, and traditional data providers like CoinMarketCap, alongside specialized crypto-native operators.

Decentralized Oracle Networks are groups of these node operators that work together to produce a single data feed. Each network has multiple independent operators sourcing the same data type (an ETH/USD price, for example) from multiple external sources. The network aggregates the responses to filter out outliers and produce a consensus value.

LINK is the network's native token. Node operators must stake LINK to participate, putting their staked tokens at risk if they provide inaccurate data. They earn LINK rewards for accurate data. This creates a direct economic incentive aligned with honest participation.

Smart contract integration happens through Chainlink price feeds that consuming protocols read from. Aave doesn't have to operate any oracle infrastructure itself. It simply reads the ETH/USD price feed from Chainlink whenever it needs to determine whether a position is at risk of liquidation.

The result is a system where the data flowing into smart contracts is not dependent on any single party. To corrupt a major Chainlink price feed, an attacker would need to compromise a majority of the node operators in that network simultaneously — and the operators have economic skin in the game to prevent that.

What Chainlink is used for

The largest use case by far is DeFi price feeds. Aave and Compound use Chainlink to determine collateralization ratios. Synthetix uses Chainlink for synthetic asset prices. MakerDAO uses Chainlink for the prices that determine when DAI vaults need to be liquidated. Hundreds of smaller protocols depend on Chainlink price feeds for their core mechanics.

Beyond price feeds, Chainlink provides:

Verifiable Random Function (VRF) — a tamper-proof source of randomness used by NFT projects for trait generation, by games for outcome randomization, and by any application that needs cryptographically guaranteed unpredictability.

Proof of Reserve — verifiable on-chain attestations that off-chain assets exist in the amounts claimed. Used by some stablecoin issuers to provide ongoing verification of reserves.

Cross-Chain Interoperability Protocol (CCIP) — a system for securely passing messages and tokens between different blockchains. Launched in 2023, CCIP is positioned as a replacement for less-secure bridge architectures that have historically been the largest single source of hacks in crypto. Chainlink ran successful tests of CCIP with the interbank messaging service SWIFT, suggesting the infrastructure is being positioned for traditional-finance integration.

Functions — a service letting smart contracts trigger arbitrary off-chain computation (calling APIs, running code) with on-chain verification of the result.

What happens when oracles fail

Oracle failures have caused some of the largest single-event losses in DeFi history.

In November 2020, Compound experienced what was effectively an oracle attack on DAI. The DAI price on Coinbase briefly spiked due to a thin-market trading anomaly. Compound used Coinbase as one of its price sources at the time. The brief spike triggered $90 million in liquidations of DAI positions that should not have been liquidated. The error was the protocol's pricing source not being adequately diversified, not Chainlink itself.

In several other instances, smaller DeFi protocols that relied on insufficiently-diversified price oracles (sometimes just one DEX pool, sometimes a small number of feeds) have been drained by manipulating the pool prices momentarily to trigger profitable liquidations or trades. These attacks are now sophisticated enough that they form their own category of DeFi exploit.

Chainlink's economic security and diversified data sources have made it the most-trusted oracle in the space, but it has not been immune to issues. The protocol has continuously evolved its security model in response to new attack patterns.

The competitive landscape

Chainlink is by far the largest oracle network, with the broadest blockchain support and the deepest integrations across DeFi. But it is not the only option.

Pyth Network specializes in high-fidelity financial market data sourced directly from major trading firms, with a focus on low-latency price feeds for derivatives platforms. Pyth has grown rapidly since 2022 and is now competitive with Chainlink in some specific markets.

Band Protocol is a smaller oracle network with multi-chain support, focused on cost efficiency and customization for specific use cases.

API3 takes a different approach — instead of running independent oracle networks, API3 promotes "first-party oracles" where data providers run their own nodes directly, eliminating the middleman.

Each has tradeoffs. Chainlink remains dominant in part because of network effects (most protocols already integrate it) and in part because of its consistent operational track record. The category is competitive but Chainlink is the default choice for most new DeFi protocols.

The practical takeaway

When you evaluate any DeFi protocol, look at what it uses for price data. The answer tells you more about the protocol's risk profile than almost anything else.

Protocols that use multiple independent oracle sources and aggregate them are structurally more secure than those relying on a single source. Protocols that use Chainlink benefit from the largest economic security of any oracle network. Protocols that rely on a single DEX pool for prices are vulnerable to brief manipulation attacks. Protocols that source prices from a centralized exchange API can be affected by the exchange's own thin-market anomalies.

Oracles are infrastructure most users never directly interact with. They are also the most likely single source of failure in any DeFi system that depends on external data. Understanding this layer is what separates participants who can evaluate protocol risk from participants who can only read headlines.