1. What LINK actually does

LINK has three distinct functions inside the network. Conflating them is the source of most confusion:

Related misconception corrections: LINK is only a payment token ↗ LINK has no role in network security ↗

2. Where the revenue comes from

Chainlink is frequently described as "just price feeds." In practice the network operates several distinct fee-generating services. Each is a real product with real users; their relative size varies, and Chainlink does not publish a full audited revenue breakdown, so treat any precise revenue figure you see — anywhere — with appropriate caution.

• Data Feeds ↗ — the original service: aggregated market data delivered onchain, securing a large share of DeFi collateral.
• Data Streams ↗ — low-latency, pull-based market data for applications needing sub-second updates. Subscription and per-use fees.
• CCIP ↗ — fees for cross-chain messages and token transfers. The service institutions use for cross-ledger settlement.
• Automation ↗ — fees for reliably triggering smart contract functions without centralized bots.
• Functions / compute ↗ — fees for offchain computation delivered with cryptographic guarantees.
• SVR (Smart Value Recapture) ↗ — redirects MEV value that would otherwise be captured by block builders back toward the oracle network and the application. Converts a leak into a revenue source.
• BUILD ↗ — partner protocols commit a portion of their own token supply in exchange for oracle services and support, creating token flows to stakers that are not funded by new LINK issuance.

Each service links to its own page — what it does, who uses it, and the evidence. Also see: CRE ↗ Proof of Reserve ↗ ACE ↗ CCID ↗ VRF ↗

3. Payment Abstraction and the Chainlink Reserve — how fees become LINK

This is the mechanism that connects network usage to the token, and it is the part most often missed. It works in two steps.

Step one — Payment Abstraction ↗. Users do not have to hold LINK to pay for Chainlink services. They can pay in gas tokens, stablecoins, or (for enterprises) offchain in ordinary fiat invoices. Those payments are then programmatically converted into LINK onchain, using Chainlink services and decentralized exchange infrastructure. The friction of "you must acquire our token to use our product" is removed for the customer, while the fees still end up denominated in LINK.

Step two — the Chainlink Reserve ↗. Launched on 7 August 2025, the Reserve is an onchain reserve that accumulates the LINK produced by that conversion process. It is funded from both onchain service fees and offchain enterprise revenue. Chainlink states that no withdrawals are expected for multiple years, so accumulated LINK is effectively locked away as it builds.

The split is not a single fixed ratio across all services — it varies by service and by whether that service is secured by staking. The clearest published example is SVR: recaptured MEV is first split between the application and the Chainlink Network (Aave’s launch terms were 65% Aave / 35% Chainlink), and the Chainlink portion is then directed onward. Chainlink states that 50% of fees from staking-secured SVR services — originally earmarked to cover node operator rewards — is now planned to help fund the Reserve, with the remainder going to node operators and community stakers.

Why this matters structurally: rewards funded by fees behave very differently from rewards funded by new issuance. One depends on customers using the product; the other dilutes existing holders. This design is an explicit attempt to shift from the second to the first.

4. How staking works, in plain terms

Full staking page ↗

Staking in Chainlink is not the same as staking on a proof-of-stake blockchain. There are no blocks being produced and no transactions being ordered. Instead:

• Node operators and community participants lock LINK into a staking pool.
• That locked LINK acts as a bond: it can be slashed if the oracle service fails to meet its commitments.
• Rewards come from a combination of protocol emissions and, increasingly, fees generated by network services.
• The larger the value secured by an oracle network, the more meaningful the bond needs to be — which ties security spending to real usage rather than to speculation.

The economically important idea is alignment: the party providing the data has money at risk if the data is wrong. That is what a bond does, and it is why staking is a security mechanism first and a reward mechanism second.

5. Emissions and dilution — the honest version

Staking rewards have historically been funded substantially by protocol emissions from reserves rather than purely by service fees. This is a real dilution consideration and it deserves to be stated plainly rather than waved away.

The stated design direction is that as network usage grows, fee revenue becomes a larger share of what funds rewards, and emissions a smaller one. Payment Abstraction and the Chainlink Reserve are the infrastructure built for that transition, converting real customer payments into LINK rather than minting it. Programs like BUILD add further non-dilutive flows — value arriving from partner protocols rather than from new issuance.

6. Why institutional adoption is the variable that matters

The tokenomics only function at scale if the services are actually used at scale. That is why this site spends its effort documenting verified institutional adoption rather than modelling prices: adoption is the input, and it is observable. Everything downstream of it is inference.

Whether a central bank runs a settlement pilot, whether a market infrastructure moves from proof-of-concept to production, whether a bank uses CCIP for real transfers — these are checkable facts, and they are the honest place to focus.

Browse the evidence vault ↗ See the adoption map ↗

7. What this page deliberately does not do

No price targets. No supply-and-demand models projecting a value. No "if X% of the market adopts, then LINK is worth Y" arithmetic. Those calculations are trivial to construct and consistently wrong, and publishing them would undermine the point of a research vault.

Nothing on this page is investment advice. Read the methodology and disclosure ↗ for how this site handles evidence, uncertainty, and conflicts of interest.