Defilama

Defilama is a dashboard for comparing assets held by DeFi protocols

Defilama is a dashboard that shows how much cryptocurrency applications hold and compares that value across blockchains. That dollar total is called total value locked, or TVL, while DeFi means financial software running through blockchain contracts. The service combines open-source adapters, token prices and protocol classifications to present chain and protocol histories. Researchers use it to compare Aave, Lido, Uniswap and other systems without assembling every contract balance themselves.

Do not read every TVL move as new deposits

A Defilama TVL chart is a dollar-valued balance series, so a falling line does not automatically mean users withdrew assets.

TVL combines token quantities with token prices. If Ether loses dollar value while an Aave market holds the same quantity, the displayed TVL falls even though its contract balances do not. The 1-day change therefore answers a valuation question, while USD inflows aim to isolate balance changes between consecutive days. Extend the review to the 7-day window, inspect the underlying token mix and note any methodology event attached to the chart. Those checks separate price translation from deposits, withdrawals and adapter changes before the movement enters a report. For the closely related topic, continue with Defilama explainer.

Defilama beside Dune, Token Terminal and CoinGecko

Defilama is a cross-chain market dashboard, while Dune, Token Terminal and CoinGecko organize research around different data models.

Use Defilama when the first decision concerns comparable TVL, chain totals, yield pools, stablecoin supply, fees or exchange volume across many protocols. Dune starts with queryable blockchain datasets and custom SQL dashboards, which suits questions that require bespoke event logic. Token Terminal emphasizes standardized financial and usage metrics for protocols. CoinGecko begins with token markets, prices and supply data. Nansen adds labelled-wallet and address-level analysis. The overlap is real, yet each tool chooses a different unit of comparison.

A strong workflow starts with the aggregator whose data model matches the question, then uses another model when the claim crosses categories. The related steps are listed in Defilama choosing comparison.

The first research pass from chain to protocol

The first Defilama research pass is a narrowing workflow that moves from a chain total to one protocol’s stated methodology.

Set the chain frame

Begin with the Chains view and choose the network relevant to the question, such as Ethereum, Solana, Arbitrum or Base. Compare the same 30-day window across candidates, then widen to 1 year when a short move needs context. Chain TVL sums included protocol TVL, so it describes the tracked application layer rather than every asset associated with the network.

Narrow the protocol set

Move from the chain into one category, such as lending, decentralized exchanges or liquid staking. Compare Aave with other lending protocols or Uniswap with other exchanges instead of mixing business models. Open the protocol record, separate deployed versions where the interface exposes them and check which chains contribute to the parent total.

Read the methodology

The methodology states which contracts, vaults or balance sources the adapter counts. Match that scope to the research claim before exporting a number. Document whether the figure counts supplied assets, liquidity-pool reserves or another contract state. A protocol total built from supplied collateral answers a different question from a fee series built from user payments, even when both appear on the same product page.

What does Defilama measure?

Put another way, Defilama measures several on-chain market categories, with TVL serving as one metric rather than a universal score for protocol quality.

TVL and inflows

Protocol TVL is the dollar value of coins held in that protocol’s contracts, while chain TVL sums included protocols on that chain. USD inflows compare asset balance changes between consecutive days and value those changes in dollars. A 24-hour TVL percentage describes the marked value; an inflow figure targets movement in asset quantities. This distinction matters most when volatile assets dominate the contract balance.

Fees and revenue

Fees record the value users pay to operate through a protocol. Revenue is the portion retained by the protocol, its treasury or token holders after supply-side payments. Defilama also separates token-holder revenue where supported. Read the 7-day and 30-day fields together because one busy day distorts a short annualized comparison.

Volumes, yields and stablecoins

DEX volume sums trades through decentralized exchanges, while yield records describe pools and their annual percentage yields. Stablecoin views track circulating supply and peg data across chains. A 1-year field supplies broader context for fees or revenue, but a displayed APY remains a live pool observation rather than a fixed return. The comparison changes when incentives alter the pool’s reward component.

Reading TVL alongside inflows, fees and revenue

A multi-metric Defilama view is a diagnostic frame that separates asset-price movement, capital flow, user payments and protocol income.

Worked example: all changing inputs in this paragraph are hypothetical. A lending protocol holds 1 000 ETH on day one at a price of $2 000 per ETH, producing $2 000 000 of TVL. On day two it still holds 1 000 ETH, while the price falls to $1 600. TVL becomes $1 600 000, a 20% decline, but net USD inflows from balance changes equal $0. The concrete result for the stated case is a $400 000 TVL decline caused entirely by price.

Apply the same sequence to real records: note the token quantities represented by the adapter, separate inflow from valuation and compare fees with revenue. High TVL with little fee activity describes a different operating profile from lower TVL with sustained usage. The interpretation changes when the protocol’s accounting unit, product category or adapter scope changes.

Coverage differences across chains and protocol versions

Protocol coverage is an adapter map, so researchers must check chain support and product-version support as separate questions inside Defilama.

Names alone are weak join keys for exported data. Ethereum mainnet uses chain ID 1, Optimism uses 10, BNB Smart Chain uses 56, Polygon PoS uses 137, Base uses 8453, Arbitrum One uses 42161 and Avalanche C-Chain uses 43114. Keep those identifiers beside readable chain names when combining records from different tools. At protocol level, Aave V2 and V3 or Uniswap V2, V3 and V4 represent different contract systems, deployment footprints and fee logic. A parent total may group versions, while a version record may expose a narrower slice. Compare like with like when the version boundary changes the counted contracts.

Filters that turn a ranking into a research view

For most users, Defilama filters are analytical controls that turn a broad leaderboard into a comparable set of chains, categories or protocols.

Lock the same currency unit, category and time horizon before sorting. A 1-day move highlights recent change, 7 days smooth one session, 30 days supplies a monthly frame and 1 year tests whether the recent direction persists. Sort only after filters match. Otherwise, a large multi-chain protocol, a single-chain deployment and a chain-wide aggregate occupy one list while answering different questions.

Use category filters to keep lending, liquid staking and decentralized exchanges in separate cohorts. Then pair TVL with the metric tied to the operating mechanism: active loans for Aave, DEX volume for Uniswap or staking-related data for Lido. A ranking becomes useful when every row shares a scope and the chosen metric matches the product.

Open-source adapters under the dashboard

An open-source adapter is a calculation program that reads blockchain state, normalizes token balances and returns metric values to Defilama.

Balance collection

Adapter code identifies contracts and calls the functions needed to recover balances at a chosen block or timestamp. On EVM chains, an address contains 20 bytes and appears as 40 hexadecimal digits after the 2-character 0x prefix, making the full text form 42 characters. Exact addresses matter because similarly named tokens and protocol versions can use different contracts.

Unit and price normalization

Unit normalization converts raw integers into token quantities before Defilama applies a dollar price. Ethereum WETH uses 18 decimals, USDC uses 6 and WBTC uses 8. One ether equals 10 to the power of 18 wei. Reading the declared precision prevents a 6-decimal stablecoin balance from being treated like an 18-decimal token balance. That conversion runs before aggregation, so a precision error propagates through protocol, category and chain views without changing the raw balance.

Aggregation and history

Returned balances feed protocol, category and chain aggregates after the methodology applies exclusions and de-duplication rules. Historical execution at an earlier timestamp rebuilds a series from past chain state when the adapter and RPC path support it. This mechanism links visible charts to inspectable code, while data quality changes when contracts, pricing coverage or chain access changes.

Exclusions that protect comparability

Regardless of the route taken, Defilama exclusions are accounting boundaries that keep native staking, bridge custody and internal receipt-token loops from inflating comparable chain totals.

Native token staking that secures a chain does not enter chain TVL under the core methodology. Liquid staking protocols receive separate tracking, yet their values stay outside chain TVL by default. Bridge projects retain their own TVL, while that custody does not contribute to origin or destination chain TVL. Smart-contract wallet balances such as Safe or Argent also stay outside the count. Within one protocol, a deposited asset and the receipt token created from it count once rather than twice. For lending, active loans sit outside TVL by default, which limits circular borrowing from magnifying the same supplied capital.

These boundaries matter whenever two dashboards use the same label but assign different accounting scope.

Data access for spreadsheets and code

Broadly, Defilama data access is an export layer that moves dashboard metrics into spreadsheets, scripts and repeatable monitoring workflows.

Spreadsheet functions cover TVL, borrowed amounts, fees, revenue, holder revenue, volume, perpetual volume, market capitalization, stablecoins, fully diluted valuation, outstanding fully diluted valuation and price. Documented fee and volume timeframes include 24 hours, 7 days, 30 days and all history. Historical calls accept a start date and an optional end date, while yield functions return pool records with chain, project, TVL, APY and token fields.

Use a 4-digit year, 2-digit month and 2-digit day for date-shaped inputs, then store the metric name beside each export. The top-pools spreadsheet function uses 20 rows as its documented default, so an explicit limit makes downstream reports easier to reproduce. API output still inherits the adapter methodology; code access changes transport and scale, not the meaning of the field.

Limits, verification and ongoing maintenance

A Defilama research routine is a maintained process that records definitions, timestamps and adapter scope before any comparison enters a report.

Save the protocol slug, chain, category, selected timeframe and extraction time with every observation. Reopen the methodology when a chart breaks its prior pattern, a product deploys new contracts or a parent listing gains another version. The chart and the code may update on different schedules, so a reproducible note preserves what the analyst actually compared.

For contributor maintenance, TVL changes belong in the open adapter repository, while fees, revenue and volume use the dimensions adapter system. A merged TVL listing may take up to 24 hours to reach the interface. Retest historical points when a balance source or pricing route changes. That interval is a publication step, not a measurement window, so timestamp the visible record after it appears.

Finish by checking that the chain identity, version scope, token precision and accounting exclusions still match the question. TVL gives scale, fees show user payments, revenue shows protocol retention and inflows separate asset movement from price. The final interpretation changes whenever one of those definitions changes.