What is The Graph?
👋 New to this? Just start reading at the top — it begins in plain English and gets more detailed as you scroll. Jump to any level:
🟢 The simple version
Plain English — no jargon. Start here.
The Graph is a decentralised indexing protocol for blockchain data — it organises on-chain events from Ethereum and other blockchains into searchable "subgraphs" so that DeFi applications can query blockchain history efficiently without running their own full nodes.
The blockchain data problem
Blockchains store data in a format optimised for security and immutability — not for querying. If you want to find "all Uniswap trades for a specific token pair in the last 24 hours," querying Ethereum directly requires scanning millions of blocks. This is impractically slow for user-facing applications. Before The Graph, DeFi protocols either ran their own centralised databases (defeating decentralisation) or accepted slow query times.
The Graph solves this by indexing blockchain events into structured, queryable databases. A "subgraph" is a set of rules defining what events to capture from which smart contracts and how to organise them. The Graph's network of indexers processes these subgraphs and serves GraphQL queries — so when you load Uniswap's interface and it shows you your recent trades, it is almost certainly querying a subgraph via The Graph.
The four participants
Indexers run nodes that index blockchain data according to subgraph specifications and serve queries, earning GRT for query fees and indexing rewards. Curators stake GRT on specific subgraphs to signal that a subgraph is high-quality and worth indexing — curators earn a portion of fees from the subgraphs they curate. Delegators stake GRT with indexers (similar to Livepeer's delegator model) to earn a share of indexer rewards without running infrastructure. Subgraph developers publish the indexing specifications that define what data to capture.
What GRT is used for
GRT is used for staking (indexers must stake GRT as a security deposit, slashable for misbehaviour), curation (curators bond GRT on subgraphs), delegation (delegators stake GRT with indexers), and paying query fees. GRT has a fixed supply of 10 billion with inflation for network rewards, partially offset by query fee burns.
Is it legal in India?
Yes. GRT is a VDA under Indian law. 30% tax and 1% TDS apply. See India regulation.
🟡 A bit more detail
For when you want to go a little deeper.
Subgraphs — how blockchain data gets organised
A subgraph is defined by three files: a manifest (specifying which contracts to watch and which events to capture), a schema (defining the GraphQL data model — what entities exist and their fields), and mappings (AssemblyScript code that transforms raw events into the schema's entities). When an indexer processes a subgraph, it replays all historical blockchain events matching the manifest's specifications and builds a database according to the mappings.
Once indexed, applications query the subgraph via GraphQL — a flexible query language that allows asking exactly for the data you need in a single request. For example: "Give me the top 10 Uniswap v3 pools by volume in the last 7 days, with their current fee tier and TVL." This query would be impractical to answer from raw blockchain data but takes milliseconds against an indexed subgraph.
Virtually every major DeFi protocol uses The Graph. Uniswap's analytics interface (info.uniswap.org) is powered by a Uniswap subgraph. Aave, Compound, Curve, Synthetix, and hundreds of other protocols rely on subgraphs for their dashboards, historical data, and in-app analytics. The Graph processes billions of queries monthly. Over 30,000 subgraphs have been deployed. Applications can also build on the hosted service (Graph's managed centralized indexing) before migrating to the decentralised network.
The curation signal mechanism
Curators stake GRT on subgraphs using a bonding curve — the first curator to stake on a subgraph gets the lowest entry price; subsequent curators pay progressively more for the same share of curation signals. When a subgraph receives many query fees, curators earn proportionally. When curators unstake, they receive GRT minus a burn (preventing rapid signal manipulation). This bonding curve mechanism creates economic incentives for curators to identify high-quality, high-usage subgraphs early — rewarding accurate curation.
Key Graph metrics: queries served per day, number of active subgraphs, total GRT staked (across indexers, curators, delegators), and protocol fee revenue. The Graph's usage is directly correlated with DeFi activity — more DeFi usage = more subgraph queries = more GRT demand. The transition from hosted service to fully decentralised network is the key development to track. Live data: CoinGecko · The Graph Explorer.
🟣 The full technical picture
For the technically curious.
The Graph whitepaper: the query market
The Graph whitepaper describes a "query market" where indexers compete to serve queries at the best price and quality. Consumers (DeFi applications) pay query fees in GRT. Indexers set their own fee rates. Gateways route queries to indexers based on price, speed, and reliability. The market mechanism creates price discovery for blockchain data queries — expensive data becomes cheaper as competition increases, and high-value data commands premium prices that attract more indexers.
Source: The Graph documentation. thegraph.com/docs · The Graph whitepaper: The Graph Whitepaper (GitHub)
Slashing and dispute resolution
Indexers stake GRT as a security deposit. If an indexer provides incorrect data (serving wrong query results), they can be slashed — losing a portion of their stake. Slashing is triggered by dispute resolution: any participant can challenge an indexer's attestation (a cryptographic signature over a query result). The Arbitration Charter defines what constitutes valid grounds for slashing. Arbiters (initially controlled by the Graph Foundation, transitioning to decentralised arbitration) evaluate disputes. This slashing mechanism creates economic incentives for indexers to serve accurate data — incorrect data risks losing significantly more than the query fee earned.
The Graph's subgraph manifest format uses a YAML/JSON schema specifying: dataSources (smart contract addresses and ABI), eventHandlers (which contract events trigger mappings), and blockHandlers (arbitrary block processing). Mappings are written in AssemblyScript (a TypeScript subset that compiles to WebAssembly) and executed by the indexer's node. Each event handler receives the decoded event parameters and can read/write to a key-value store that represents the subgraph's entity state. The entity state is queryable via GraphQL. Subgraph deployment is done via IPFS — the manifest, schema, and compiled WASM mappings are uploaded to IPFS, and their content hash is the subgraph's identifier on-chain.
Key facts
- Function: Decentralised blockchain data indexing
- Query language: GraphQL
- Participants: Indexers, Curators, Delegators, Developers
- GRT supply: 10 billion (fixed total, with inflation for rewards)
- Chains supported: Ethereum, Arbitrum, Polygon, Solana, +40 more
- Usage: Uniswap, Aave, Compound, Curve, 30,000+ subgraphs
- Founded: 2018 (Yaniv Tal, Brandon Ramirez, Jannis Pohlmann)