The Graph

The Graph is a suite of blockchain data infrastructure products that extract, process and deliver structured blockchain data across 60+ networks (the docs site also references 57 networks in its network listing). Its stated purpose is to solve the difficulty of reading blockchain data: chains have no relational structure or query language, contract state lives in low-level storage slots and event logs, historical queries require archive nodes, and application data is fragmented across many contracts. Rather than each developer building bespoke indexing infrastructure, The Graph offers a modular protocol in which different data services address different classes of data problem.

The core products are Subgraphs (custom open APIs that define which contracts, events and entities to index and serve the transformed data over GraphQL, aimed at dapp frontends, DeFi dashboards, NFT marketplaces and governance UIs); Substreams (parallelized streaming indexing for low-latency, high-throughput use cases such as live liquidity and price feeds, trading and liquidation events, analytics pipelines and large backfills); and Amp, a blockchain-native database that turns onchain activity into enterprise-ready datasets with SQL access, audit trails, compliance and verifiability features for regulated and institutional users. Supporting components documented publicly include Graph Node, which indexes blockchain data and serves it over GraphQL, and Firehose, which extracts blockchain data into flat files to speed sync times.

The protocol is operated as a decentralized network described as a data market with distinct participant roles: Indexers who run nodes and serve queries, Curators who signal which subgraphs are useful and accurate, Delegators who stake in support of Indexers, and subgraph developers. The Graph publicly claims outcomes including elimination of self-run data servers, 60–98% lower monthly cost versus running infrastructure, and 99.99%+ uptime through a globally distributed contributor set.

Founding story

The three founders — Yaniv Tal, Jannis Pohlmann and Brandon Ramirez — had previously worked together at startups including Workflo, TapSavvy and MuleSoft. They began exploring Ethereum in early 2017 and concluded that immature tooling and protocols made building dapps difficult; making open data more accessible became their focus. The team built a first prototype in late 2017, seeking a productive developer experience for writing indexing logic that could be securely operated on a decentralized network. The Graph launched in June 2018. [7]

Business model

The Graph operates a decentralized data marketplace in which independent Indexers run nodes and process queries in an open market, Curators signal on useful subgraphs, and Delegators support Indexers; network fees are earned by members of this distributed community rather than by a single operator. Developers consume data through subgraph queries, Substreams packages, and API keys, and enterprise/institutional access is offered through Amp with a direct "contact team" sales motion. Earlier in its history the project ran a centralized hosted service before transitioning to the decentralized network. [0][2][3][5]

Sources describe query fees flowing through an open marketplace of Indexers rather than a conventional company revenue line; The Graph markets itself as a "competitive data market" where developers pay for query service and network fees are distributed to participants. No pricing or company revenue figures are disclosed in the sources. [0][5]

Traction

As of early 2026 the company reports over 1.27 trillion queries served to more than 75,000 projects, 60+ supported networks, and 100+ Indexer nodes on the website (the docs cite 50+ independent Indexer nodes). Earlier milestones: seven launch partners in January 2019 (Dharma, Compound, Uniswap, ENS, Origin, Decentraland, Livepeer); over 1,500 subgraphs deployed by mid-2020; a hosted service processing roughly 50 million queries per day and over 750 million queries in May 2020, up 45% month over month; and, as of October 2020, over 200 Indexer nodes in testnet and more than 2,000 Curators in the Curator Program. In January 2022 the network supported data from 26 blockchains. [0][3][5][6][7]

Latest developments

The website's lead announcement is Amp, described as a blockchain-native database that turns complex onchain data into structured, verifiable insights, positioned for enterprise and regulated use cases with SQL access, compliance features and audit trails. Documentation states that as of early 2026 the network had served over 1.27 trillion queries to more than 75,000 projects across 60+ chains, and public GitHub repositories show active commits through August 2026, including a Rust rewrite of indexer-service with TAP payments, a rewards eligibility oracle for indexer performance, and AI agent skills for subgraph development. [0][3][4]