Airnode is open-source oracle software developed by the API3 project that lets the owners of a web application programming interface connect it directly to blockchains. Instead of relying on independent node operators to fetch and relay data, the data provider runs a lightweight Airnode itself, so smart contracts receive information signed by the original source. API3 calls this a "first-party" oracle.
Airnode sits in the blockchain oracle market, which supplies off-chain data such as prices, weather or sports results to on-chain applications. Its approach contrasts with third-party oracle networks such as Chainlink, where independent operators gather data from many sources. API3 is governed by a decentralized autonomous organization whose members hold the API3 token.
Airnode in plain terms
Blockchains cannot look things up on the internet by themselves. If a lending app needs the price of a coin, someone has to bring that number in. Many oracle systems use couriers who collect the number from a website and deliver it. Airnode asks the website owner to deliver it personally, with their own signature on the envelope.
The idea is that cutting out the middle courier removes one place where data could be delayed, changed or charged for twice. The trade-off is that you now depend heavily on the honesty and reliability of each data provider, so many providers are usually combined.
History and API3
API3 was launched in 2020 by Burak Benligiray, Heikki Vänttinen and Saša Milić, who had earlier worked on oracle projects, and published a whitepaper that year arguing for first-party oracles. The project sold API3 tokens and set up the API3 DAO to manage its treasury and direction. Airnode was its first major product, followed by data feeds called dAPIs that aggregate signed data from several providers. In later years API3 put more emphasis on these feeds and on capturing "oracle extractable value", the profit available to whoever triggers actions such as liquidations right after a price update, and returning part of it to the applications that use its data.
How Airnode works
- Serverless deployment. Airnode is designed to run as serverless functions on cloud platforms, so a provider can deploy it with a configuration file and leave it running with little maintenance.
- Request and response. A smart contract makes a request on-chain; the provider's Airnode detects it, calls the provider's API and writes a signed response back.
- Signed data feeds. Providers can also sign data off-chain, and anyone can push the signed values on-chain, which supports regularly updated price feeds.
- Chain compatibility. Airnode supports Ethereum and other networks that run the Ethereum Virtual Machine.
Because the response is signed with the provider's key, users can check which source supplied each value.
Uses
Airnode-based data is used mainly in decentralized finance, where lending markets and derivatives platforms need reliable price feeds to value collateral and trigger forced liquidations. Other possible uses include insurance contracts that pay out on weather data, games that need random numbers and applications that need traditional market or sports data. API3 has offered coverage arrangements funded by staked tokens, intended to compensate users if a data feed malfunctions, subject to governance approval.
The API3 token and governance
Holders of the API3 governance token can stake it in the DAO's pool to vote on proposals, such as funding teams and setting fees, and to earn rewards. Staked tokens have also been designed to back the coverage service, so stakers take on some of the risk of feed failures. As with many DAOs, voting power is concentrated among large holders, and much of the practical development is carried out by teams funded by the DAO.
Risks and criticism
First-party oracles remove intermediaries but concentrate trust in data providers. If a provider's API is wrong, hacked or goes offline, its Airnode will faithfully deliver bad data or nothing at all, so aggregating several providers is essential. Running in a cloud environment creates dependence on large cloud companies. Oracle manipulation remains a major source of DeFi losses across the industry, and attackers may try to profit by moving prices on the sources that feeds rely on. Critics also note that the number of high-quality data providers willing to run their own nodes has been limited, which affects how widely the first-party model has been adopted.