What is RPC provider?

A company that runs nodes and sells access to them, so applications can read chain data and broadcast transactions without operating infrastructure.

Nearly every wallet, explorer, and application depends on one, which concentrates a network that is decentralized at the consensus layer into a handful of access points.

Running your own node means downloading and validating the chain yourself, which costs disk, bandwidth, and attention. Almost nobody does it, so almost everybody reads the chain through somebody else's node. That is a reasonable trade for most people, and it is worth knowing what it costs you.

How it works

1. The provider operates nodes, usually including archive node instances that retain full historical state, and exposes them over an HTTP or websocket endpoint.
2. Your wallet points at that endpoint. When it shows a balance, estimates a fee, or simulates a transaction, it is asking the provider, not the chain.
3. Broadcasting goes the same way. A signed transaction is handed to the provider, which relays it to the network.
4. The provider sees the requests. Addresses queried, transactions broadcast, and the IP address they came from all pass through one company's infrastructure.

Two consequences follow. The first is availability: an outage at a large provider makes wallets appear broken on a chain that is producing blocks normally. The second is privacy: a provider can associate the addresses you query with each other and with your network location, which links wallets you may have kept separate.

Your wallet's private keys never touch the provider. It cannot spend anything. What it can do is see what you ask about, and return a wrong answer if it wanted to.

Example

Illustrative. You hold four wallets in one browser extension and take care never to transact between them. The extension uses one default endpoint for all four.

Every balance refresh queries all four addresses from the same connection, so the provider can trivially associate them. Nothing was broadcast on chain and no transaction links them, yet the association exists in one company's logs.

Changing the endpoint splits that, and running your own node removes it. Most wallets expose a custom network setting where the endpoint can be replaced in under a minute, which is why endpoint choice is the usual privacy lever as well as the usual fix when a wallet stops loading.

Why it matters when you buy

When a wallet shows a wrong balance or refuses to send after a purchase, the endpoint is a more likely cause than the chain, and switching it is faster than anything else you might try. How to use a block explorer covers checking the chain independently, and the chain pages cover the networks themselves.

Related terms

node — what the provider runs for you; full node — validating the chain yourself; archive node — retains full history; light client — verifies without full data; block explorer — an independent second view; indexer — structured query layer above raw access.

Questions

Can my provider steal my crypto?

No. Signing happens in your wallet and the keys never leave it. A malicious or broken provider can show wrong data or fail to broadcast, which is a different problem and still worth guarding against.

Why did my wallet stop showing balances?

Most often an endpoint problem rather than a chain problem. Check a block explorer to see whether the network is producing blocks, then switch the endpoint in your wallet's network settings.

Should I run my own node?

It is the strongest option for privacy and verification and it costs disk, bandwidth, and maintenance. For most people, choosing an endpoint deliberately and verifying important transactions independently is the reasonable middle.