Skip to content
Get started

SOL · Account model · since 2020

Solana blockchain explorers: decoding is the whole job

A Solana transaction is not one action. It is a bundle of instructions hitting several programs, any of which may have failed while the others succeeded — and reading that bundle is the entire value an explorer adds.

Updated 17 September 2026 · 5 explorers compared · How we test
Try
Ticker
SOL
Launched
2020
Block time
about 400 milliseconds per slot
State model
Account
Live lookup here
Yes

Account: Each address holds a running balance that transactions debit and credit.

The best Solana explorers

Solana has fewer credible explorers than Ethereum and the differences between them are larger. The decision is mostly between richer decoding and the ability to point the tool at your own node.

  1. 1

    Solscan

    3.4 /5

    The default. Instruction-level decoding, SPL token resolution and the best NFT handling on the chain. Closed, and it cannot be pointed at your own RPC.

    Free tier, key required Full review
  2. 2

    The official, open-source option, and the only one that reads from any RPC endpoint you give it — including a validator on your laptop. Less decoding, more verifiable.

    Free, no key Open source Full review
  3. 3

    3xpl

    4.1 /5

    Fast, keyless API, works without JavaScript. The right choice for confirming a transaction exists without waiting for a heavy interface.

    Free, no key No-JS Full review
  4. 4

    Arkham

    3.4 /5

    Entity attribution across Solana and twenty-four other chains. For following where funds went rather than confirming they moved.

    Paid only Full review
  5. 5

    Dune

    3.6 /5

    SQL over indexed Solana data. The tool when the question is "how many" rather than "what happened here".

    Free tier, key required

One transaction, many instructions

On Bitcoin a transaction moves value from inputs to outputs. On Ethereum it is a call from one address to another. Both are simple enough that a naive explorer produces something readable.

Solana is structurally different. A transaction carries a list of instructions, each naming a program to invoke and the accounts that program may read or write. A single swap might contain eight instructions across four programs: create a token account, approve a delegate, invoke the exchange program, which internally invokes two liquidity pools, then close the temporary account. All atomic, all one transaction.

Shown raw, that is a wall of base58 addresses and byte arrays — technically complete and practically useless. The entire value of a Solana explorer is in decoding: knowing which program is the SPL Token program, that this instruction is a transfer, that the mint has six decimals so the raw integer means what it means.

Solscan maintains that decoding layer across the major programs and keeps it current as new ones appear. That is unglamorous, continuous work and it is why it is the default.

Failed transactions are normal here

This surprises everyone arriving from another chain, so it is worth stating plainly: on Solana, failed transactions are recorded on chain, cost a fee, and are common.

During busy periods a meaningful share of submitted transactions fail. Slippage exceeded. An account changed state between simulation and execution. The compute budget was exhausted. Your wallet may say "transaction failed" and leave you to work out which.

The important reassurance is atomicity. A Solana transaction either completes entirely or does nothing at all. There is no partial state where your tokens left but never arrived. If it failed, the only thing that happened was the fee — which is small.

Solscan shows which instruction failed, which program returned the error, and the error code with a decoded meaning where it has one. For the common cases that turns an opaque failure into something actionable: custom program error 6001 on a swap is usually slippage, and compute budget exceeded means the transaction needed more compute units than it requested.

Where your tokens actually live

Everything on Solana is an account: wallets, token balances, program code and program state. Each has an owner — the program permitted to modify it — and a lamport balance that must meet the rent-exemption minimum.

Your wallet address does not hold your USDC. It owns a separate associated token account, which holds it. One wallet with five different tokens has five token accounts, each a distinct on-chain address with its own rent-exempt balance.

So an address page has to show two different things: the SOL balance held by the wallet itself, and the token balances held by accounts it owns. Solscan handles this clearly. A generic multi-chain explorer frequently does not, which is why a Solana balance can appear to be missing on one tool and present on another.

The rent mechanic follows from the same design. Accounts must hold a minimum SOL balance to persist, so creating a token account costs a small amount — returned when you close it. It is a deposit, not a charge, though it does mean a wallet with no SOL cannot receive a token it has never held.

Why the official explorer still matters

Every commercial Solana explorer reads from infrastructure its operator controls. The official explorer reads from whatever endpoint you tell it to.

Change a dropdown and you are on devnet. Enter a URL and you are on a private cluster, a paid RPC provider, or a validator running on your own machine. For development that is not a convenience, it is the difference between having an explorer and not having one — testing a program on localnet with Solscan is impossible, because your local transactions do not exist as far as it is concerned.

It also matters for verification. Solscan decodes aggressively, and a decoder that does not recognise a program presents its best guess rather than announcing uncertainty. The official explorer shows decoded instructions alongside the raw account list and instruction data, so when two tools disagree the raw form settles it.

And it has a privacy consequence: point it at your own RPC and no third party learns what you looked at. That is the same argument Otterscan makes on Ethereum, available here without abandoning the mainstream tool.

Slots, blocks and what "confirmed" means

Solana produces a slot roughly every 400 milliseconds, and not every slot contains a block. Explorers show both slot and block height, and the two are not the same number — which confuses people comparing figures across tools.

Commitment levels are the other thing worth understanding. A transaction can be processed (seen by the node you asked), confirmed (voted on by a supermajority) or finalized (buried under enough subsequent blocks to be irreversible in practice). Explorers and RPC calls default to different levels, which is why a transaction can appear on one tool and not yet on another.

For anything consequential, finalized is the level that matters. For a wallet showing you that a swap went through, confirmed is what you are seeing, and it is almost always sufficient.

The practical consequence for exchange deposits is that Solana's thresholds are expressed in commitment level rather than confirmation count, which is a different mental model from Bitcoin's. Our guide on checking confirmations covers how the equivalent question is answered on each chain.

Solana explorers: common questions

What is the best Solana blockchain explorer?

Solscan for everyday reading and for untangling a failed transaction. The official explorer for development, custom clusters and raw data. Most people who build on Solana keep both open.

Why did my Solana transaction fail?

Commonly slippage exceeded, account state changed between simulation and execution, or the compute budget was exhausted. Failed transactions are recorded on chain and cost a fee, which surprises people arriving from other chains. The transaction is atomic — if it failed, nothing in it happened except the fee.

Why does my wallet address show no tokens?

Because Solana tokens live in separate associated token accounts that your wallet owns, not at the wallet address itself. One wallet holding five tokens has five token accounts, each a distinct on-chain address. This is the most common confusion on Solana.

Is the Solana RPC free?

The public mainnet endpoint at api.mainnet-beta.solana.com needs no key and answers, but only from a server. We tested it on 17 September 2026: the same request returns 200 from curl and 403 from a browser, because the endpoint refuses anything carrying an Origin header. If you are calling it from front-end code you want a public node that serves browsers, such as solana-rpc.publicnode.com, which is what the lookup tool on this site uses. Either way it is heavily throttled and explicitly not for production, and dedicated RPC providers require keys and offer real capacity.

What is rent on Solana?

Every account must hold a minimum SOL balance to stay alive, which is why creating a token account costs a small amount you get back when you close it. It is a storage deposit rather than a fee.

Live network data

Read from public nodes right now

BTC block height

0000000

Next-block fee

0000000

Mempool backlog

0000000

ETH block height

0000000

Next difficulty

0000000

Solana slot

0000000

Live from public node APIs · updates on load ·
Any percentage shown is Bitcoin's next mining difficulty retarget, a protocol parameter. It is not an asset price or a measure of investment performance.