Bitcoin UTXO Privacy Score Calculator

How UTXO Privacy Scoring Works

Every Bitcoin UTXO carries a privacy fingerprint. Chain surveillance firms exploit predictable transaction patterns to cluster addresses, trace payment flows, and deanonymize users. A UTXO privacy score quantifies how resistant a given unspent output is to these techniques by evaluating it across several heuristic dimensions.

Tools like OXT's Boltzmann analysis, the am-i-exposed scanner (available on Umbrel), and the Rust crate kaccy-bitcoin all produce scores on a 0–100 scale. Each tool runs dozens of heuristics, but the core scoring factors fall into six categories: address reuse, round-number detection, change output identification, timing analysis, script type consistency, and merge input clustering. The table below summarizes how each factor is weighted and what it reveals.

A score of 80–100 indicates strong privacy: the UTXO is difficult to trace or cluster. A score below 40 signals significant exposure through one or more heuristic weaknesses. Most UTXOs created by default wallet behavior score in the 40–60 range because wallets prioritize fee efficiency over privacy.

Address Reuse History

Address reuse is the single most damaging privacy practice in Bitcoin. When an address receives funds more than once, every transaction involving that address becomes trivially linkable. According to BitMEX Research, roughly half of all transaction outputs historically went to previously used addresses, though that figure has declined as wallets adopted HD wallet standards (BIP 32/44/84). As of late 2025, Glassnode data indicates approximately 4.33 million BTC still sits in addresses that have been reused, representing about 21.5% of circulating supply.

This factor is binary and deterministic: either the address has been used before or it has not. A UTXO sitting on a reused address scores zero for this component. Wallets implementing proper address rotation generate a fresh address for every receive, eliminating this vector entirely.

Round-Number Detection

Payments tend to be round numbers. When a transaction has one output of exactly 0.1 BTC and another of 0.03847291 BTC, analysts infer the round value is the payment and the irregular value is change. This heuristic extends to fiat-denominated amounts: a payment of exactly $100 worth of Bitcoin at the current exchange rate produces a precise-looking BTC amount that can still be flagged if the dollar equivalent is round.

The round-number heuristic has moderate reliability because legitimate payments can produce non-round amounts (invoices, exchange withdrawals), and change can sometimes be round by coincidence. To defeat this heuristic, some privacy-focused wallets add small random amounts to payment values. The PayJoin protocol also neutralizes this signal by combining inputs from both sender and receiver.

Change Output Identification

Identifying which output is change and which is the payment is the central goal of most chain analysis heuristics. Researchers have cataloged over twenty distinct change-detection heuristics. The most effective include:

• Script type matching: if all inputs use P2WPKH and one output is P2WPKH while the other is P2TR, the P2WPKH output is likely change
• One-time address: if one output goes to an address that has never appeared on-chain, it is likely change (wallets generate fresh change addresses)
• Largest output: in simple two-output transactions, the smaller output is often the payment and the larger is change
• Wallet-specific patterns: Bitcoin Core's coin selection runs four algorithms and picks the result with the lowest waste score, producing a recognizable transaction shape

A UTXO scores well on this factor when the transaction that created it has ambiguous outputs: same script types on both sides, similar amounts, or a structure that resembles a CoinJoin. For a deeper dive into practical coin management, see the Bitcoin privacy scoring guide.

Timing Analysis Exposure

Transaction timestamps leak behavioral metadata. Research from Meiji University demonstrated that analyzing the time distribution of a user's transactions can reveal their time zone with approximately 72% accuracy. A University of Luxembourg study showed that a moderately resourced adversary monitoring the peer-to-peer network can correlate transactions to their origin node by analyzing propagation timing, at a cost of only hundreds of dollars.

Timing exposure is scored on two axes: whether the UTXO was created during a narrow, predictable time window (indicating routine behavior), and whether the transaction was broadcast with timing patterns consistent with other transactions from the same cluster. Users can mitigate timing analysis by using delayed broadcasting, Tor-based transaction relay, or the Dandelion protocol (BIP 156), which obscures the originating node.

Script Type Consistency

Bitcoin supports multiple address types: legacy P2PKH (1...), wrapped SegWit P2SH-P2WPKH (3...), native SegWit P2WPKH (bc1q...), and Taproot P2TR (bc1p...). When a transaction's inputs and outputs use different script types, the mismatch becomes a fingerprint. If all inputs are P2WPKH but one output is P2TR, that output is almost certainly the payment destination, not change.

P2PKH addresses still secure roughly 43% of mined bitcoin by value, so mixed-type transactions remain common. The long-term fix is Taproot (P2TR) adoption: because P2TR makes multisig, single-sig, and complex scripts look identical on-chain, widespread Taproot usage reduces the information that script type differences reveal. A UTXO created in a transaction where all inputs and outputs use the same script type scores full marks on this factor.

Merge Input Clustering Risk

The common-input-ownership heuristic (CIOH) assumes all inputs in a single transaction belong to the same entity. First described in the Bitcoin whitepaper itself, this heuristic is the foundation of most chain surveillance clustering. When you consolidate UTXOs or spend from multiple addresses in one transaction, every input address becomes linked to a single identity in the analyst's graph.

CIOH has high reliability for standard transactions but breaks down for CoinJoin, PayJoin, and CoinSwap transactions where multiple parties contribute inputs. A UTXO scores poorly if it was created in a transaction with many merged inputs from addresses that appear elsewhere on-chain. The coin control feature in wallets like Sparrow and Bitcoin Core lets users manually select which UTXOs to spend, preventing accidental clustering.

Boltzmann Entropy and Transaction Ambiguity

Boltzmann analysis, developed by LaurentMT for the OXT block explorer, provides a mathematical framework for measuring transaction privacy. It computes the number of possible interpretations for a transaction's input-to-output mapping and expresses the result as entropy in bits using the formula E = log2(N), where N is the count of valid mappings.

A standard two-input, two-output transaction where both outputs are the same amount has an entropy of 1 bit (two possible interpretations). A well-structured CoinJoin with equal-value outputs can achieve entropy of 5+ bits, meaning 32 or more equally plausible interpretations. Boltzmann also produces a link probability matrix showing the likelihood that each input funded each output, making it possible to assess per-UTXO privacy within a multi-output transaction.

Improving Your UTXO Privacy Score

Each scoring factor has specific countermeasures. The following table maps weaknesses to practical remediation steps.

For a comprehensive walkthrough of coin management techniques, see the Bitcoin coin control and UTXO privacy guide and the Bitcoin privacy landscape analysis.

Privacy Scoring in Practice

The Bitcoin UTXO set contains approximately 173 million entries as of early 2025, with about 49% holding fewer than 1,000 sats. Many of these sub-dust UTXOs are inscription-related (29.6% of UTXOs by count) and carry minimal privacy risk due to their negligible value. The UTXOs that matter for privacy scoring are those with material value that users will eventually spend.

When spending a scored UTXO, its privacy characteristics transfer to the new transaction. A high-scoring UTXO spent in a sloppy transaction (reused address, mixed script types, round payment amount) loses its privacy advantage. Conversely, a low-scoring UTXO can be rehabilitated through a well-structured CoinJoin that breaks the deterministic links in its history. Privacy is not a static property: it changes with every transaction in the UTXO's lineage.

Layer 2 protocols offer a different approach to UTXO privacy. Spark moves Bitcoin and stablecoin transactions off the base layer entirely, avoiding on-chain UTXO creation for routine payments. This reduces the on-chain footprint that surveillance firms can analyze, though the privacy guarantees depend on the specific L2 architecture.

Frequently Asked Questions

What is a Bitcoin UTXO privacy score?

A UTXO privacy score is a numerical rating (typically 0–100) that estimates how resistant a specific unspent transaction output is to chain analysis heuristics. It evaluates factors like address reuse, change output detectability, script type consistency, and transaction graph exposure. Higher scores indicate the UTXO is harder to trace or link to a real-world identity. Tools like OXT's Boltzmann analyzer and the am-i-exposed scanner compute these scores using 30+ heuristics.

How does address reuse affect Bitcoin privacy?

Address reuse is the most damaging privacy practice because it creates a deterministic link between all transactions involving that address. Anyone who knows one transaction on a reused address can see every other transaction, the balance, and the full spending history. Approximately 21.5% of circulating Bitcoin supply still sits in reused addresses. Using an HD wallet with automatic address rotation eliminates this risk entirely.

What is Boltzmann analysis in Bitcoin?

Boltzmann analysis is a method developed by LaurentMT that measures transaction privacy by computing the number of valid input-to-output mappings. It expresses the result as entropy in bits: E = log2(N). A standard transaction with distinguishable outputs has 0 bits of entropy (one interpretation), while a well-constructed CoinJoin can achieve 5+ bits (32+ interpretations). The analysis also produces a link probability matrix showing how likely each input is to have funded each output.

Can I improve a UTXO's privacy score after receiving it?

Yes. The most effective method is to pass the UTXO through a CoinJoin transaction, which breaks deterministic links between inputs and outputs. Wasabi Wallet's WabiSabi protocol and JoinMarket both provide CoinJoin implementations. You can also improve privacy by spending the UTXO in a PayJoin transaction or by migrating it to a Taproot address to reduce future script-type fingerprinting. See the privacy scoring guide for step-by-step instructions.

What is the common-input-ownership heuristic?

The common-input-ownership heuristic (CIOH) assumes all inputs in a single Bitcoin transaction belong to the same entity. It was first described in the Bitcoin whitepaper and remains the foundation of address clustering by chain analysis firms. CIOH breaks down for CoinJoin and PayJoin transactions where multiple independent parties contribute inputs. Using coin control to avoid merging UTXOs from separate contexts is the primary defense.

Does Taproot improve UTXO privacy?

Taproot (P2TR) improves privacy by making single-sig, multisig, and complex script spends look identical on-chain through Schnorr signature aggregation. This eliminates the script-type fingerprinting that distinguishes P2PKH from P2SH multisig spends. However, Taproot only helps when both sender and receiver use P2TR addresses. As long as P2WPKH and P2PKH remain widely used (P2PKH still secures roughly 43% of mined bitcoin by value), mixed-type transactions continue to leak information about which output is change.

How do chain analysis firms score UTXOs?

Commercial chain analysis firms like Chainalysis and Elliptic use proprietary clustering algorithms that combine CIOH, change-detection heuristics, timing analysis, and off-chain data (exchange KYC records, subpoena responses) to build entity graphs. Their internal scoring goes beyond the open-source heuristics described here by incorporating behavioral models trained on labeled datasets. The open-source tools referenced in this guide approximate the heuristic component of commercial analysis, but they do not replicate the off-chain data enrichment that gives commercial tools their edge.

This tool is for informational purposes only and does not constitute financial or legal advice. Privacy scores are conceptual estimates based on publicly known heuristics and do not connect to live blockchain data. The weights and reliability ratings reflect commonly cited research and open-source tool implementations, not a single authoritative standard. Always verify current privacy practices and tool capabilities before making security decisions.