Responsible team: Reclaim / MarketWhale · Reviewed:

Reclaim research Scan: 9 September 2026

What an 85-wallet Solana rent scan found

A read-only scan of an archived leaderboard cohort found small rent estimates in most completed previews—and ten results we could not measure.

The main finding

Of 85 wallets attempted, 75 previews completed. Their median estimated proceeds were about 0.0505 SOL before network fees. This study measured potential empty-account closures; it did not recover SOL or measure excess withdrawals.

Most completed estimates were below 0.1 SOL.

72 previews found provisional closure candidates, three completed with no candidate accounts, and 10 failed. That is 96% provisionally eligible among the 75 completed previews, or 84.7% of all 85 attempts. Failed results are unknown, not zero.

Estimated proceeds before network fees

75 completed previews. Bars show the share of those previews in each SOL range.

  • 0 SOL3 wallets
  • Above 0, below 0.0111 wallets
  • 0.01 to below 0.146 wallets
  • 0.1 to below 114 wallets
  • 1 SOL or more1 wallet

The largest group was 46 wallets between 0.01 and 0.1 SOL. The ten failed previews are excluded from this chart.

The median describes the middle completed preview. The combined total adds only those 75 previews; neither number estimates the unmeasured wallets or Solana wallets generally.

Exact SOL estimates for all 75 completed previews, including three zero results. Proceeds exclude network fees.
MeasureAccount SOL
before deductions
Estimated proceeds
after deductions
Median0.0561364740.050522828
Combined total8.8273111597.944580111

The downloads also include percentiles, gross distributions, and a separate calculation for the 72 provisionally eligible wallets.

This is one historical cohort, with ten unknowns.

We attempted every address in the retained leaderboard snapshot: 85 rows, 85 unique addresses, no duplicates. Its embedded run date is 25 January 2026; that date is not an independently verified collection window. Addresses were sorted before scanning, with none selected or removed based on the new results.

The original article attributed the list to KOLScan, but the snapshot does not establish its original selection rules. We treat it as a convenience cohort from an archived leaderboard. Current leaderboard membership, activity, identity and ownership were not verified. Names and social profiles were not used in the analysis.

Missing oversized wallets can bias the completed sample toward fewer accounts and lower rent amounts. We did not fill in the missing values, extrapolate a full-cohort total or calculate population confidence intervals.

A preview is an estimate, not a completed recovery.

The scan queried the original Token Program and Token-2022. It selected accounts with zero parsed token amounts, preserved one empty wrapped-SOL account when present, and checked reported close authority. Gross rent means the SOL actually held by candidate accounts, summed from their lamport balances.

For this September run, estimated proceeds subtract a 6% service portion and a 4% referral portion. Both were allocated to the service in this no-referral preview model. No low-balance subsidy or service-account-rent deductions were recorded. Network fees were unknown in every preview and are excluded. These are the historical model's estimates; use a fresh quote for today's accounts and costs.

No transaction was built, simulated, signed or submitted for the study. Eligibility remained provisional: account changes, signing authority, extension rules, transaction validation and batch rounding can affect what closes and what you receive. 68 completed previews carried a Token-2022 validation warning.

This study covers empty-account closure only. It does not measure excess SOL available from accounts that remain open, or today's combined wallet-health estimate. See how excess recovery works and current fee details.

Download the data and reproduce the arithmetic.

The frozen observations and calculation script let you recompute the published totals without network or wallet access. A new live scan would observe different account state.

Download observations.json and analyze.mjs into the same directory. With Node.js 20 or later, run:

node analyze.mjs observations.json reproduced

This writes reproduced.json and reproduced.csv. They should match the published summary files byte for byte. It recomputes the recorded estimates; it does not rescan wallets.

Collection and calculation details

Collection ran on Solana mainnet-beta on 9 September 2026, from to . It used the version 1.0 preview operation on a local Reclaim server with its configured Helius RPC. Reads used confirmed commitment; successful observations span slots 445645271–445645788. This was not an atomic snapshot across wallets or necessarily across calls within a wallet.

There were 103 HTTP requests, with one retry for each of 18 wallets and 1.1 seconds between wallets. Transport failures, HTTP 429 and server errors could be retried once; account and response-limit errors were not retried. The observation JSON retains every attempt.

1 SOL equals 1,000,000,000 lamports. Calculations use exact integer lamport strings. The median is the middle sorted value, or the mean of the two middle values for an even sample; half-lamport medians are preserved. P75 and P90 use nearest rank: one-based rank ceil(p × N).

The two fee portions are separately rounded down to integer lamports. The historical model adds a 100,000-lamport deduction per candidate at wallet balances of 0.0001 SOL or less, and may deduct service-account rent. Estimated net is gross minus those deductions, floored at zero. Neither additional deduction occurred in this run. Network fees remain unknown; batch rounding can differ.

Solana's Close Account documentation describes balance and authority requirements; getTokenAccountsByOwner describes the lookup. Candidate filters and cost deductions belong to Reclaim. See the preview contract for the API.

Public rows omit wallet addresses, wallet balances, names and social handles. Removing these identifiers reduces direct identification but does not guarantee anonymity. Source, cohort, raw-run and implementation SHA-256 hashes record provenance. The files reproduce aggregate arithmetic, but cannot independently verify which wallet produced a row or the underlying RPC responses. Raw token-account RPC payloads were not retained.

Check your wallet's current estimate

These historical numbers do not predict your result. A public-address preview checks your current accounts; review its warnings and costs before considering a separate signing flow.

How to preview a wallet

Cite as: Reclaim / MarketWhale. (2026, September 9). What an 85-wallet Solana rent scan found. Study version 1.0: 85 attempted wallets, 75 completed previews. Permanent results link.

Article updated 4 October 2026. The scan date and frozen dataset remain 9 September 2026.