How much SOL is stuck in a Solana wallet? Data from 67 wallet cleanups

The short answer

Across 67 wallets that reclaimed SOL with SOLTidy between May 10 and October 5, 2026, the median wallet closed 2 token accounts and got back 0.00277 SOL after fees. The mean was much higher, 12.3 accounts and 0.0224 SOL, because a few heavy traders closed dozens or hundreds of accounts. Roughly one wallet in ten recovered 0.0778 SOL or more, and the largest single recovery was 0.3494 SOL.

Why publish this data at all?

I publish it because the honest answer to "how much SOL is stuck in my wallet" has always been "count your accounts and multiply", and I wanted to see what that looks like across real wallets. I built SOLTidy, and since May every confirmed reclaim transaction has been recorded from the chain. The per-account arithmetic lives in how much SOL you can recover from token accounts; this post adds what people actually recovered.

That earlier post says to be suspicious of any site quoting an "average" recovery figure, and that still holds. The numbers below describe 67 wallets that chose to use SOLTidy, not your wallet and not Solana as a whole. They are useful for calibrating expectations, especially the gap between the median and the mean, and useless as a prediction.

How much SOL did each wallet get back?

The median wallet got back 0.00277 SOL, the mean wallet 0.0224 SOL, and the top wallet 0.3494 SOL, all after SOLTidy's 5% fee. These are the totals for May 10 to October 5, 2026:

Measure	Value
Distinct wallets	67
Confirmed transactions	263
Token accounts closed	826
SOL returned to wallets, after the SOLTidy fee	1.4999 SOL
Period	May 10 to October 5, 2026

Per wallet, the spread looks like this:

Per wallet	Token accounts closed	SOL returned after fee
Median	2	0.00277 SOL
Mean	12.3	0.0224 SOL
90th percentile	not computed	0.0778 SOL
Largest	178	0.3494 SOL

The median of 0.00277 SOL is almost exactly two newer SPL token accounts after the fee: 2 x 0.00148844 x 0.95 is 0.00283 SOL. In other words, the typical wallet in this sample cleaned up a couple of leftover accounts and stopped there, or had little else to clean.

Why is the average eight times the median?

The mean is about eight times the median because a small number of heavy traders account for a large share of everything recovered. The largest single wallet closed 178 accounts, about 22% of all 826 closed in the sample, and recovered 0.3494 SOL, about 23% of the 1.4999 SOL total. Remove that one wallet and the mean drops from 0.0224 SOL to about 0.0174 SOL.

That shape matches how token accounts are created. Every distinct token a wallet buys or receives opens one account, and selling leaves it open. Two kinds of wallets come out of that:

• Light users. A wallet that swapped a few tokens and holds mostly SOL and a stablecoin has a handful of empty accounts at most. This is the bulk of the sample.
• Heavy traders. A wallet that traded hundreds of memecoins, or a trading bot wallet that bought every new launch, can have well over a hundred. These few wallets produce most of the SOL.

For a skewed distribution like this, the median is the better description of a typical wallet, and the mean is mainly useful for totals. Anyone quoting a single "average wallet holds X SOL" figure without saying which one is hiding most of the story.

What does recovery per token account look like after fees?

Token Account Closer transactions returned about 0.0018 SOL per closed account after the 5% fee, across 815 accounts in 256 transactions, or a little over 3 accounts per transaction. For comparison, the after-fee value of one empty SPL token account is 0.00141 SOL if it was opened after Solana lowered rent and 0.00194 SOL if it was opened before.

The observed 0.0018 SOL sits toward the top of that range, and I would not read too much into it. Two things push it up: some of the accounts were opened at the old rent rate, and the closer also withdraws spare rent from older accounts users keep and closes pump.fun volume accumulator accounts, which add SOL without adding a closed token account to the count. So it is best read as "SOL returned per token account closed, including extras", not as the deposit of a single account.

By tool, the sample looks like this:

Tool	Transactions	Token accounts closed	SOL returned after fee
Token Account Closer	256	815	1.4877 SOL
Sell Dust for SOL	6	10	0.0108 SOL
NFT Burner	1	1	about 0.0014 SOL

Almost everything recovered so far came from closing token accounts. Dust selling and NFT burning barely appear yet, partly because they are permanent actions that people pick item by item, and partly because the sample is small.

How was this data collected?

Every figure comes from confirmed Solana transactions read back from the chain, not from anything the browser reports. The process works like this:

1. When a SOLTidy transaction confirms, the page sends only its signature to the server.
2. The server fetches that transaction from a Solana RPC node and rejects it unless it succeeded and the fee payer sent a payment to the SOLTidy fee wallet.
3. From the chain data it records the paying wallet, the number of token-program CloseAccount instructions, the SOLTidy fee, the network fee and the wallet's balance change.
4. Each signature is stored once, so a transaction cannot be counted twice.

"SOL returned" is the sum of each transaction's positive balance change for the wallet, after the SOLTidy fee; Solana's network fee of 5,000 lamports per signature is negligible next to it. Scans are not recorded at all, since they sign nothing. Transactions done through the home page cleanup are attributed to the underlying tool that built them. The live totals are on the Solana wallet cleanup page and grow as new transactions confirm.

What are the limitations of this data?

The biggest limitation is the sample: 67 wallets over five months is small, early and self-selected. Specifically:

• Self-selection. These are people who found a reclaim tool and chose to use it. Wallets with nothing to reclaim never show up, and neither do wallets cleaned with other tools, wallet menus or the spl-token CLI.
• What users chose, not what was there. The amounts are what users decided to reclaim, not everything locked in their wallets. Someone might close ten accounts and leave forty open, or skip NFTs and stake accounts entirely.
• Possible undercounting. A transaction is only recorded if the page reported its signature. If a tab closed at the wrong moment, the transaction happened on chain but may be missing here.
• Token-account focus. "Accounts closed" counts token-program closes only. Spare-rent withdrawals and pump.fun account closes add SOL but not to that count.
• No dollar figures. I report SOL only, because the dollar value depends on the day you read this.

I will update these figures as the sample grows and note the new date at the top of the post. With a few hundred wallets the percentiles will mean considerably more than they do now.

How do you check how much SOL is stuck in your own wallet?

Run a read-only scan: it counts your actual accounts and shows the SOL in each, which beats any average. The quickest way is the free wallet cleanup scan on the home page:

1. Connect your wallet. Phantom, Solflare, Backpack and other Solana wallets work.
2. Let the scan read your token accounts, NFTs, stake accounts, pump.fun rewards and token approvals. Scanning signs nothing and costs nothing.
3. Read the per-source totals, which are shown after fees: empty token accounts, spare rent, pump.fun accounts, deactivated stake, NFTs and dust.
4. Decide. If the figure is a couple of thousandths of a SOL, like the median wallet here, closing a few accounts from your wallet or the spl-token CLI for network fees only is a perfectly good choice. If it is closer to the top of the table, batching saves a lot of clicking.

If you only want empty token accounts, the Token Account Closer does that one job and charges 5% of the SOL it recovers. To turn any account size into a deposit yourself, the Solana rent calculator uses the live network rate, and the rent-exempt minimum by account type lists the common ones. And if your balance looks low for other reasons, why your Solana balance is lower than expected covers the other places SOL goes.

Bottom line

In SOLTidy's first 67 on-chain-verified cleanups, the typical wallet got back very little, 0.00277 SOL from 2 accounts, while the mean of 0.0224 SOL was carried by a few heavy traders, one of whom recovered 0.3494 SOL alone. Per closed token account, users kept about 0.0018 SOL after the fee. The sample is small and self-selected, so treat it as a sense of scale, not a forecast, and run a free read-only scan to see your own number. I will keep updating these figures as more wallets come through.