The Graph Subgraphs & Blockchain Data

What The Graph is

The Graph is a family of blockchain-data products rather than one interchangeable API. Subgraphs define and index application-specific data into GraphQL endpoints. Substreams are high-throughput streaming pipelines. Graph Node is the open-source indexing runtime for teams that need self-hosting or unsupported EVM networks. The Gateway handles hosted query access, API keys, and billing.

Best fit

• Applications that need a stable, typed query layer over repeatable onchain data.
• Teams willing to define a schema, mappings, deployment process, and freshness monitoring instead of relying on an instant universal API.
• Data engineers who need Substreams or self-hosted Graph Node for custom throughput, network, or infrastructure requirements.

Check before choosing

• The first 100,000 monthly queries are currently free and additional queries cost $2 per 100,000. Use a production API key with domain restrictions and a monthly spending limit; do not expose an unrestricted key in client code.
• Querying and indexing are separate economic activities. Paying more query fees cannot make an under-indexed Subgraph available or current. Verify Indexer support and incentives, or operate Graph Node when the hosted path does not meet the requirement.
• Deploying to Subgraph Studio is not the same as publishing to the decentralized network. Production teams need an explicit publish and versioning process.
• Network support differs between Subgraphs, Substreams, and Firehose. The public network covers more than 60 chains, but product support is not identical on every network; unsupported EVM networks may require self-hosting.
• Check deployment version, number of Indexers, latest synced block, timestamp, and indexing errors before treating results as current. New networks and triggers can expose untested errors or synchronization lag.
• The graph-node repository remains actively maintained, with release v0.45.0 published in August 2026. Repository activity is a durability signal, not an uptime or freshness guarantee for a specific Subgraph.

Decision summary

Choose The Graph when an application-specific indexed query layer is worth the schema, deployment, and monitoring work. Prefer a hosted data provider when you need immediate cross-chain coverage without owning indexing definitions. Before launch, test one representative Subgraph end to end, restrict its key, set a spending cap, measure block lag and indexing errors, and document a fallback for stale or unavailable data.