dApp (decentralized application)
An application whose backend logic runs on smart contracts on a public blockchain rather than centralized servers.
A dApp typically combines a frontend (hosted on traditional or decentralized infrastructure) with backend smart contracts. The user connects a wallet, signs transactions, and interacts directly with on-chain code — there's no central authority that can freeze accounts, censor users, or change rules unilaterally.
Major dApp categories: DEXes (Uniswap, Curve), lending (Aave, Compound), NFT marketplaces (OpenSea, Blur), gaming (Axie Infinity), social (Lens, Farcaster), and governance (Snapshot, Tally). The smart contract layer provides settlement; the frontend is interchangeable.
Pure dApp purity varies. A dApp where all logic is on-chain is rare — most "dApps" use centralized services for indexing, frontend hosting, or even some backend logic. The degree of decentralization is a spectrum, and most successful "dApps" are hybrid in practice.
dApps are the user-facing surface of crypto. Their adoption determines whether smart-contract platforms have real demand beyond speculation.
How CryptoRadar24 tracks it
CryptoRadar24 reports on dApp activity through TVL, transaction volume, and ecosystem growth metrics.
Related terms
FAQ
Is every crypto app a dApp?
No. Centralized exchanges (Binance, Coinbase) are crypto apps but not dApps — their core logic runs on private servers, not public smart contracts.
Can a dApp be banned?
The smart contract typically cannot be removed (immutable on-chain). But specific frontends can be blocked, app stores can refuse listings, and developers can be sanctioned. Determined users can still access via direct contract interaction.
Are dApps free?
Using one means paying gas for transactions. The dApp itself is usually free to access (the smart contract has no listing fee), though some have protocol fees on usage.
Why are dApps slower than regular apps?
Every state-changing action requires an on-chain transaction — wallet signing, mempool wait, block confirmation. Read-only operations (browsing, viewing) can be fast. Writes are bottlenecked by chain throughput.