Linea RPC Nodes & Endpoints

Linea RPC URLs you can use right now

Third-party endpoints, free and shared. Linea's campaign-driven traffic means these are usually comfortable and occasionally, during a big incentive push, exactly when you need them most, not.

import { JsonRpcProvider } from "ethers";

const provider = new JsonRpcProvider("https://rpc.linea.build");
const blockNumber = await provider.getBlockNumber();
console.log("Linea block:", blockNumber);

Where shared Linea endpoints stop being enough

Linea's MetaMask distribution means its public endpoints absorb some of the most bursty retail traffic in the EVM world: a quest campaign, an airdrop checker, or a featured dApp can multiply request volume in an afternoon. Shared endpoints spread that pain across every tenant, so your app slows down because someone else's campaign went viral. If your Linea product has traffic spikes of its own, you are exposed twice.

A dedicated Linea RPC node isolates you from the crowd. Your quest backend, indexer, or wallet API gets the node's full capacity during exactly the windows when usage peaks, and eth_getLogs ranges stay a configuration choice rather than a plan restriction. On BLAZED.sh the workload itself runs on the node host in a container, so reads travel a local socket instead of the public internet.

Three kinds of finality, and which one your app should believe

A transaction on Linea passes through three distinct states, and treating them as one is the source of most confusing bug reports on zk rollups. The sequencer accepts and orders it first, which is what makes the UI feel instant. The prover then generates a validity proof for the batch containing it, which takes materially longer than block production. Only when that proof is verified on Ethereum is the transaction final in the sense that it cannot be undone without breaking L1 itself.

For a wallet balance or a feed, the sequencer's view is the right one to display and the delay is invisible. For anything that moves value off the chain, it is not: a bridge, an exchange crediting a deposit, or a payout job reading the sequencer head is acting on state that is ordered but not yet proven. The gap is minutes to hours depending on batch cadence, not seconds, which is long enough to matter and short enough that testing on a quiet day will never surface it. Decide per read path which finality you require, and query the block tag that matches it rather than defaulting to latest everywhere.

Linea RPC latency is a distance problem

Linea behaves like mainnet from a tooling perspective, and it inherits the same RPC physics: a remote gateway adds tens to low hundreds of milliseconds of public-internet round trip to every call, before any queueing. The structural fix is proximity. On our live Ethereum product, co-located containers see sub-10ms calls over a local socket, and dedicated Linea nodes will ship on the same architecture.

What dedicated Linea node hosting will cost

BLAZED.sh prices node access in per-request credits granted monthly by your subscription, with a deliberately narrow weight spread: most RPC methods cost a single credit, and only a handful of heavy operations cost more. Dedicated Linea nodes will launch on the same credit model, with the waitlist deciding rollout priority.

Retail-facing Linea apps are read-dominated, and reads cost one credit each, so scaling from a quiet week to a campaign week changes the count, not the price per call.

Perfect for

What is Linea?

Linea aims for full EVM-equivalence using a zk proof system, so existing Ethereum contracts and tooling generally work without modification, a meaningful difference from zkEVMs that require special compilation. It benefits from Consensys' ecosystem reach through MetaMask and developer tooling, which drives real end-user traffic to Linea RPC endpoints. Batches are proven and verified on Ethereum for validity-based security, and gas remains ETH-denominated. Because the developer experience mirrors mainnet, teams often extend an existing Ethereum integration to Linea; the practical scaling limit then becomes RPC capacity for logs, traces, and subscriptions under growing usage, which a dedicated node addresses directly.

Linea RPC: frequently asked questions