> For the complete Linea documentation index, see [llms.txt](/llms.txt). > Agents can fetch this page as Markdown at [https://docs.linea.build/protocol/architecture/rpc-services.md](https://docs.linea.build/protocol/architecture/rpc-services.md). # RPC services RPC services expose the JSON-RPC interfaces that applications, users, and internal services use to read network data and submit transactions. Any [Lineth](/protocol/reference/zero-knowledge-glossary#lineth)**Lineth** (Formerly the Linea Stack) The open-source ZK-rollup stack, codebase, and technical protocol that's the foundation of Linea Mainnet. Operators can deploy this stack to launch their own Ethereum-compatible L2 or L3 networks. node can serve them, publicly or only within the operator's network, and RPC nodes are typically placed behind load balancers and [access control](/stack/deployment/access-control). Two things distinguish one RPC node from another: which methods it serves, and how much state history it retains. Serving RPC is independent of a node's consensus role. ## Method availability Standard Ethereum JSON-RPC namespaces are available on any execution client. The `linea_` namespace is implemented by Linea Besu plugins, so a node serves those methods only if it runs Linea Besu with the relevant plugins enabled. A node running Geth, Erigon, Nethermind, or upstream Besu serves standard Ethereum methods but not the `linea_` namespace. ## State retention By default, a node keeps current state and recent history. This is all it needs to verify blocks and follow the chain, and it suits low-latency queries against recent blocks and current state, along with high-throughput transaction propagation. Access to this recent window is sometimes called **near-head** access. A node in this configuration can't reconstruct state at older blocks. Retaining more history costs proportionally more storage, so operators provision it only where historical queries are needed. ### Archive nodes An archive node retains state at every block from genesis, alongside the blocks and transactions themselves, so it can reconstruct the chain state at any past block height. It trades minimal latency and modest storage requirements for completeness, and acts as the source of truth for historical state, supporting auditing, forensics, analytics, and historical consistency checks. ## Why run your own RPC Public RPC endpoints are convenient for development, testing, and low-volume usage. As usage grows, or as requirements around reliability, trust, or data access get stricter, a dedicated RPC node becomes more valuable: - **Reliability**: Public endpoints are shared infrastructure, subject to rate limits, request prioritization, and maintenance windows outside your control. Your own node lets you size resources and manage uptime to your application's needs. - **Performance**: A self-hosted node can be tuned for one workload, whether that's low-latency access to recent state or high-throughput historical queries, without contention from other users. - **Correctness**: Querying a public endpoint means trusting a third party to serve accurate and complete data. Your own node validates chain data directly from the protocol. - **Historical data**: Public endpoints often restrict archive access because of its storage cost. Your own [archive node](#archive-nodes) has no such limits. - **Operational control**: A self-hosted node can sit inside a private network behind load balancers and RBAC, which matters for backend services, indexers, and internal tooling that need predictable access patterns and tighter security boundaries. ## See also - See [Deployment architecture](/stack/deployment) for how RPC nodes fit into an operator's footprint. - On Linea Mainnet, most users reach RPC services through their [wallet](/network/how-to/connect-wallet), and developers either use a [node provider](/network/build/tools/node-providers) or [run their own node](/network/how-to/run-a-node).