Optimism (OP Mainnet) RPC Nodes & Endpoints

OP Mainnet RPC URLs you can use right now

Third-party endpoints, free and shared. Fine for reads; the limits appear when an indexer backfills a wide block range or a dashboard subscribes for thousands of users at once.

import { JsonRpcProvider } from "ethers";

const provider = new JsonRpcProvider("https://mainnet.optimism.io");
const blockNumber = await provider.getBlockNumber();
console.log("Optimism block:", blockNumber);

Where shared OP Mainnet endpoints run out

Public OP Mainnet endpoints are fine for a wallet checking a balance, but Superchain-era workloads look different: portfolio apps read Optimism alongside Base and other OP-Stack chains, indexers backfill months of eth_getLogs history, and governance tooling replays proposal and vote events. Those are sustained, wide queries, precisely the shape shared endpoints throttle first, because one tenant's backfill degrades everyone else's reads.

A dedicated Optimism RPC node removes the contention. Your backfill can run at whatever rate the hardware sustains, subscriptions stay connected, and the block-range window is a client setting rather than a plan tier. BLAZED.sh takes it further by running your indexer or service in a container on the node host, so even the heaviest log query travels a local socket instead of the internet.

One OP Stack, several chains, and no shared state

Optimism is both a chain and a codebase, and the second meaning causes real integration mistakes. OP Mainnet, Base, Zora, and a growing list of others all run the OP Stack, which is why the same ethers or viem code talks to all of them after a URL swap. What the shared stack does not give them is shared state. Each chain has its own sequencer, its own block numbering, its own contract deployments, and its own bridge. An address holding a balance on OP Mainnet holds nothing on Base by default, and a contract at a given address on one is unrelated to the address on another.

The practical consequence is that Superchain support is a per-chain integration, not one integration with a dropdown. Every chain needs its own RPC endpoint, its own indexed history, and its own set of deployment addresses in config. Teams that model it as a single network with a chain ID parameter tend to find out during a cross-chain bug hunt, when a lookup silently returns the right shape of answer from the wrong chain.

OP Mainnet RPC latency is a distance problem

Optimism itself gives you two-second blocks; a remote RPC gateway adds a public-internet round trip of tens to low hundreds of milliseconds on top of every call, plus queueing when the gateway is busy. The architectural answer is to shorten the path, not to shop for a faster gateway. On our live Ethereum product, co-located containers see sub-10ms RPC because calls never leave the machine, and dedicated Optimism nodes will use the same design.

What dedicated Optimism 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 Optimism nodes will launch on the same credit model, with the waitlist deciding rollout priority.

Indexing OP Mainnet is mostly eth_getLogs and plain reads, which are single-credit calls, so a backfill budget can be estimated by simply counting the block ranges you plan to walk.

Perfect for

What is Optimism?

Optimism pioneered the optimistic-rollup approach and open-sourced the OP Stack that now powers Base, Blast, and dozens of other chains. Because those chains share a codebase, tooling and RPC semantics carry over cleanly, and an integration written for OP Mainnet usually ports to any Superchain member with a URL swap. Optimism posts transaction data to Ethereum and finalises through fault proofs, and its ecosystem skews toward DeFi, public-goods funding, and consumer apps. As with any rollup, the chain itself is rarely the constraint; sustained eth_getLogs ranges, trace calls, and websocket subscriptions are what expose weak RPC infrastructure, which is where a dedicated node earns its keep.

Optimism RPC: frequently asked questions