Verdict
Nansen's labelling turns anonymous addresses into named actors, which is genuinely valuable for research and genuinely dangerous as a trading strategy. The Smart Money framing sells a signal the data cannot really support.
- Best for
- Wallet-level research across EVM chains
- Cost
- Free tier; paid plans from ~$99/month
What works
- The largest labelled address database in the industry
- Fund and treasury flow tracking is excellent for research
- Token holder analysis surfaces concentration risk in seconds
- Coverage spans most major EVM chains plus Solana and Bitcoin
What does not
- Smart Money labels imply a predictive edge the data does not establish
- Paid plans start around $99/month and rise steeply from there
- Labels can be stale, and sophisticated actors split across unlabelled wallets
- Copying flows after they appear means trading behind the actor you are copying
Blockchains are public and pseudonymous, which means everything is visible and nothing is identified. Nansen's core product closes that gap: millions of addresses labelled with who they belong to — exchanges, funds, protocol treasuries, market makers, bridges, known individuals.
That database is the moat. It is expensive to build, expensive to maintain, and it makes on-chain data far more useful than it is raw. Everything else in the product is an interface on top of it.
Where it earns its price
Treasury and fund tracking: watching where a named entity's assets move, in something close to real time.
Holder concentration: a token's holder list with labels tells you in seconds whether the float is real or whether four wallets control it.
Protocol forensics: following funds after an exploit, across chains and mixers, with the counterparties named where they can be.
Flow research: which categories of actor accumulated or distributed over a period, which is a genuinely hard question to answer any other way.
For research, diligence and post-incident analysis this is the best tool in the category and there is no close second.
The Smart Money problem
Some wallets are labelled Smart Money, and the interface invites you to watch what they buy. It is compelling, it is the feature that drives subscriptions, and it is the weakest part of the product.
The label is assigned on past performance. So was every fund that subsequently blew up.
You see the transaction after it confirms. The entry you are copying is already reflected in the price you will pay.
A sophisticated actor who knows they are labelled can split execution across unlabelled wallets — or use the labelled one deliberately to create exactly the flow you are watching.
You see the buy. You do not see the hedge, the OTC leg, or the broader position this trade is one part of.
Following labelled flow is not front-running. It is back-running, and the person you are following knows you are there.
Coverage
Strongest on Ethereum and the major EVM chains, with Solana and Bitcoin support added and improving. Labels on newer chains are sparser, and label freshness varies — an address labelled as a fund two years ago may have changed hands, been delegated, or simply been retired.
There is no staleness indicator on most labels, which is the single improvement that would most increase their usefulness.
Pricing
The free tier gives a taste of the interface. Paid plans start around $99 a month and the professional tier runs an order of magnitude higher.
At $99, the question is whether wallet-level research changes your decisions enough to justify roughly $1,200 a year. For a researcher, a fund or anyone doing diligence professionally, easily. For someone with a five-figure portfolio making a handful of decisions a year, almost certainly not — and the temptation to justify the subscription by trading more is a real cost on top of the fee.
Verdict
Where the labels come from
Labels are built from a combination of public disclosure, on-chain heuristics, exchange deposit patterns and manual research. The methodology is described in general terms rather than published in detail, which is defensible commercially and does mean you cannot audit a label you are relying on.
In practice the labels on major entities are reliable and the labels on smaller ones vary. Before building an argument on a single labelled address, check whether the label is consistent with the address's actual behaviour — a fund label on a wallet that only ever interacts with one contract is probably stale.
The best use of it
Diligence on a token before buying: who holds it, how concentrated it is, whether early wallets are still there.
Post-incident forensics, where labelled counterparties turn a list of hashes into a narrative.
Understanding aggregate behaviour over months rather than reacting to individual transactions.
Verifying claims — a project asserting institutional backing can be checked against who actually holds the token.
Every item on that list is research. None of them is a trade signal, and the product is at its best when you use it for the former.
Score: 7.4. An outstanding research tool with a marketing layer that sells a trading signal the underlying data does not support. Buy it to understand what happened. Do not buy it to be told what to buy.