Monero for Darknet Buyers: XMR Wallets, Fees, and Privacy Basics

Privacy on Monero is not a setting. Every transfer hides the sender, the recipient, and the amount, whether the user wanted that or not. That uniformity is why darknet marketplaces started adding XMR next to Bitcoin, and why a few dropped Bitcoin entirely. It is also why a jammed mempool in March 2024 could freeze withdrawals without anyone cracking the cryptography. This is a research note on wallets, fees, and what the protocol actually conceals. It is not an access guide, and it does not point at live markets.

Why the payment rail changed

Bitcoin’s public ledger made the old workflow brittle. A deposit address, a change output, and a later cash-out could be stitched into a graph. TRM Labs has described the resulting shift: marketplaces adopted Monero alongside Bitcoin, and some moved to Monero only, because the chain hides the source of funds and the identities on both sides of a trade.

That popularity had a cost at the on-ramp. Binance and OKX delisted Monero. KYC venues got harder to use. Volume moved toward decentralized exchanges and instant swap desks that skip identity checks. TRM’s read is blunt. Casual holders lost easy access. People already operating in high-risk categories mostly did not stop.

I treat both sides of that story with the same skepticism. Law enforcement likes a clean narrative about “untraceable coins.” Market admins like a clean narrative about perfect anonymity. Neither matches the protocol, and neither survived March 2024 cleanly.

The three things a public chain would leak

On Bitcoin, a transaction permanently records who sent, who received, and how much moved. Monero’s design, as laid out on monero.how, answers each leak with a separate mechanism. They are not duplicates. Drop one and a different kind of surveillance still works.

Ring signatures and the sender

When an output is spent, the wallet bundles the real one with decoy outputs already on the chain. An observer sees that one member of the ring was spent. They do not see which member. That is plausible deniability, not a proof that analysis is impossible. Over many transactions it wrecks the simple “follow the coins” graph people still expect from Bitcoin.

The limit is practical. Decoys only help if they are plausible. A flooding event, a badly chosen wallet, or a user who immediately consolidates everything into one output can shrink the story those decoys tell. The math does not save sloppy key hygiene.

Stealth addresses and the recipient

The address you publish is not written onto the chain as the destination. Each incoming transfer creates a one-time output. Only the recipient, using their private view key, can recognize and later spend it. Outsiders cannot match that output to the address printed on a vendor page or a swap ticket.

Subaddresses matter here for a boring reason. Reusing one receiving address across every counterparty does not print your identity on-chain the way a Bitcoin address reuse does. It still links those payments inside your own wallet, and it links them to anyone who watched you hand out that same string. A fresh subaddress per desk, per app, per counterparty is the habit monerica’s wallet notes keep repeating. It is the right habit.

RingCT and the amount

Ring Confidential Transactions mask the value. The network can still check that inputs and outputs balance and that nobody minted coins. Observers cannot sort transfers by size, which is how a lot of “private” coins still get clustered. Amount privacy is why a 2 XMR test and a much larger follow-up do not advertise themselves by denomination alone. They can still advertise themselves by timing. More on that below.

The network layer is a separate problem

Ring signatures, stealth addresses, and RingCT are on-chain. They say nothing about which IP first announced the transaction. Monero uses Dandelion++ for that. The transaction is relayed along a randomized path before it is broadcast widely, which makes it harder to tie the broadcast to the originating address.

Harder is not the same as solved. Dandelion++ sits outside the on-chain protocol. A bad exit, a wallet that broadcasts in the clear from a home connection, or a swap desk that logs the deposit address next to your other coin’s withdrawal still joins data the chain refused to join. I would not describe Dandelion++ as cover for sloppy network use. It reduces one surveillance category. It does not erase the others.

March 2024, when the chain got loud

In March 2024 the network was flooded with transactions. Congestion got severe. Some users could not send, and some services could not pay out withdrawals. TRM notes the obvious speculation: was this vandalism, a stress test, or an attempt to deanonymize people by forcing weird behavior under load?

Nobody in the public write-ups has a clean attribution, so I will not invent one. What the episode did show is operational. Monero’s privacy does not guarantee liveness. A buyer who needed a payout during that window learned that “private” and “available” are different properties. The congestion also pushed the community toward further hardening, which is the useful outcome. It should have killed the idea that the protocol is finished.

Wallets: who holds the seed

Custodial balances are not Monero in any sense that matters. If a desk or a market wallet can freeze the coins, the ring signatures are protecting their keys, not yours. Non-custodial is the baseline.

The wallets that keep showing up in practical no-KYC write-ups are Cake Wallet on Android and iOS, the Monero-only wallet at Monero.com, and Monerujo on Android. I have no loyalty to any of them. Cake is the one people actually open on a phone. Monerujo is the Android specialist. A Monero-only app has a smaller attack surface than a multi-coin wallet that also stores the BTC you just swapped. That is a tradeoff, not a slogan. Multi-coin convenience is how deposit addresses from two chains end up in the same backup file.

Generate the seed yourself. Write it down. Keep it offline. Do not screenshot it into a cloud photo library. Confirm you can restore before you move a meaningful balance. A wallet you cannot restore is a donation to the void, and Monero has no support desk that can reverse it.

Integrated addresses and legacy payment IDs are mostly a leftover. Most modern services do not require a payment ID if you give them a subaddress. If a form still demands one, treat that as a sign the counterparty is old or sloppy, and read their instructions twice before sending.

After the delistings, the on-ramps got worse

Two routes dominate the public guides, and both are annoying in different ways.

The fast path is a crypto-to-XMR swap. You already hold BTC, USDT, or something equally liquid. An instant desk such as Quickex or PegasusSwap, or a comparison pass through Trocador, quotes a rate. You paste a fresh subaddress, send the deposit, and wait for XMR. Aggregators people actually check include Trocador, Swapzone, CoinoSwap, CypherGoat, and Intercambio. The quote you liked ten minutes ago is not a promise.

The slower path is peer-to-peer. Hodl Hodl and other P2P boards list offers. You filter on payment method, rate, and reputation, then follow that platform’s escrow. Some no-KYC P2P desks also run a Tor site. Atomic swaps sit beside both routes: eigenwallet for Bitcoin against Monero, BasicSwap if you want to run the software yourself, RetoSwap as a self-hosted P2P network. Atomic is the cleanest trust model if you already have BTC and you can tolerate the setup. It is also the one most people will not finish on the first afternoon.

Buying without KYC is lawful in many jurisdictions. It is not lawful everywhere, and reporting rules do not vanish because a desk skipped the selfie. That is your problem to check, not the protocol’s. “No-KYC” is also a marketing line. The same guides warn that risk systems can still demand information on large or unusual flows. I believe that warning more than I believe the badge on the homepage.

Fees, spreads, and the number that matters

There is no single Monero fee to quote, and anyone listing one is guessing. What you actually pay on a swap is a stack:

• The network fee on the coin you send in, which is often BTC or ETH and can dwarf the XMR side.
• An XMR payout fee where the desk charges one.
• The spread inside the quoted rate.
• A minimum size. Under that minimum the route simply will not start.

Compare two or three aggregators on the effective “you send / you get” numbers, not on a headline percentage. Check the ETA. Then send a small test. Wait for real confirmations in your own wallet. Save the order id and the transaction id before you scale up. Address typos and wrong networks are how people lose funds. USDT on TRON is not USDT on Ethereum. The chain will not refund a mismatch, and the desk usually will not either.

On a congested network the fee market and the wait both change. March 2024 was the public reminder. If a service’s withdrawal queue stalls, your RingCT balance is still private and still stuck. Plan for that. Do not treat an instant-swap ETA as a settlement guarantee.

Habits that leak more than the chain does

A few failure modes show up again and again in the practical guides, and they are more useful than another diagram of a ring signature.

• One subaddress for everyone. Convenient, and it ties those counterparties together in your records and in theirs.
• Skipping the test transfer. The first time you discover a wrong chain or a wallet that wants a payment ID should not be the large transfer.
• IP habits at the desk. Some instant services discourage Tor or a VPN because their fraud checks fail closed. If you get blocked, the documented fallback is a Tor-friendlier aggregator or a P2P trade, not a louder retry from the same session.
• Keeping the pre-swap coin and the XMR in one mental “wallet” while the logs still join them. The swap desk saw your BTC deposit and the XMR address you pasted. On-chain privacy starts after that handoff, not before.
• Timing. A test, then the main amount, then an immediate spend, is a pattern. Amounts are hidden. Clocks are not.

None of this is a trick for hiding a crime. It is the boring residue of how the tools actually behave. Researchers reading seizure write-ups should assume investigators start at the exchange, the swap ticket, and the device, not at a magical break of RingCT.

What I would not rely on

I would not rely on a market’s built-in wallet. Admins get arrested, exit, or “upgrade” the hot wallet. Your view key does not help if they held the spend key.

I would not rely on a single instant desk. Spreads move, minimums change, and a no-KYC label can flip into a document request when the amount looks interesting.

I would not talk about Monero as if analysis has been disproven forever. Mandatory privacy raises the cost of graph work a lot, which is why the coin displaced Bitcoin on some markets. Flooding, delistings, sloppy swaps, and reused addresses are all still in the record. The protocol did its job in 2024 by staying up and staying private under junk traffic. Users who needed a withdrawal that week learned the rest of the system did not.