Arkham now puts full wallet history, cross-chain tracing and risk scores behind a paid Enterprise tier. I followed a real $26 million theft with two Bitquery tools, then lined up five more options.

Arkham’s free plan now shows only the last 100 transactions of any wallet. To see the rest, you need Arkham Enterprise. That paid tier also holds risk scores, case files and tracing from one chain to another. Arkham launched it on 20 October 2026, and its announcement lists no price.

A busy wallet can make 100 transactions in an afternoon. So if you chase stolen funds or check wallets for a compliance team, the free plan now stops early. If that is you, you need an Arkham Intelligence alternative, at least for part of the work.

I wanted a fair test, so I took one real theft and followed it to the end. On 8 January 2026, an attacker took 8,535 ETH from Truebit, an Ethereum project with its own token. That was about $26.4 million at the time. Within 53 hours, all of it had gone into Tornado Cash, a mixer that pools many people’s coins so the trail breaks.

I ran that trail through both Bitquery tools myself. For the other five, I say what each does best and where it stops.

The short version

• Bitquery MoneyFlow (lite.bitquery.io): a free, Arkham-style flow chart in your browser. Paste a wallet, pick one of 17 chains, and follow the money one step at a time.
• Bitquery MCP: for people who work in Claude, ChatGPT or Cursor. Ask in plain English, and the AI pulls live transfers and labels from 19 chains, plus risk scores for Ethereum, Bitcoin and TRON.
• Nansen: for trading research and “smart money” wallets.
• Chainalysis Reactor and TRM Labs: what police and large compliance teams buy. Prices come by quote.
• Dune: for people who write SQL and want dashboards.
• Etherscan: free, and fine for checking one wallet or one transfer at a time.

Why people look for an Arkham Intelligence alternative

Arkham Intelligence is a blockchain analytics platform, and it built its name on labels. A label is a name attached to a wallet address, such as “Binance hot wallet”. Arkham says it holds 3.9 billion “attributions”, its word for these labels. Those labels stay free, along with its graph tool, called Visualizer, and its block explorer.

Here is what sits behind Arkham Enterprise on 20 October:

• the full transaction history of a wallet, with filters (free users see the 100 most recent transactions)
• Tracer 2.0, which follows money across bridges, the services that move coins between chains
• the counterparties panel, which shows who a wallet deals with most
• risk scores and case management
• creating new custom labels

Arkham does give free Enterprise access to some independent investigators and public-interest groups through a grant program. Everyone else has to ask for a price.

The API is gated too, with one exception. You apply through a form, and once Arkham approves you, paid plans start at $100, billed in monthly credits. Arkham also hosts an MCP server, so Claude, Cursor or Codex can query its data, but it needs that API plan or a trial. The exception is for AI agents: about 90 read endpoints sold one call at a time, at $0.20 per credit in USDC, with no account.

I get why Arkham moved full history behind a paywall. Serving full history for every wallet costs money. Still, Arkham keeps one big edge. If your main question is who owns a wallet, its free labels are hard to beat, and nothing on this list replaces them one for one. The tools below compete on history, access and price instead.

The test case: the $26 million Truebit hack

Truebit sold its TRU token through its own sale program on Ethereum, written in 2021. A math bug in that old code made the price of new TRU come out as zero. The attacker minted TRU for free, sold it back for ETH, and repeated the trick several times in one transaction. Security firm SlowMist put the loss at $26.44 million.

Here is the trail, as the two Bitquery tools showed it:

1. 8 January, 16:02 UTC. 8,535 ETH lands in the attacker’s wallet, 0x6C8E鈥�2b50 (transaction).
2. Same day. The wallet splits the money: 4,267 ETH to one new wallet, 4,001 ETH to another. A third payment of 268 ETH follows on 10 January.
3. 8 to 10 January. The 4,267 ETH pile hops through three more new wallets. The 4,001 ETH pile makes one hop.
4. 10 January. Two final wallets feed Tornado Cash in round amounts of 100, 10, 1 and 0.1 ETH. By 20:52 UTC, 8,535 ETH has gone in, the same amount that was stolen.

Any tracing tool should answer three questions about a trail like this. Where did the money come from? How did it split? Where did it end up?

1. Bitquery MoneyFlow: visual crypto tracing in your browser

MoneyFlow is Bitquery’s browser tool for tracing money between wallets. It lives at lite.bitquery.io, and it is free: you only need to sign in with a Bitquery account.

Starting a trace takes four steps:

1. Paste an address.
2. Pick the chain. On 29 October the menu listed 17: Bitcoin, Ethereum, BNB Smart Chain, Polygon, Arbitrum, Base, Optimism, TRON, XRP Ledger, Cardano, Litecoin, Dogecoin, Dash, Algorand, Filecoin, Arc and Robinhood Chain.
3. Pick the coin or token. MoneyFlow lists every currency the wallet has touched, with the date it last moved.
4. Set a date range if you want one, then press Create MoneyFlow.

The result is a flow chart. Money coming in sits on the left, money going out sits on the right, and the width of each band shows the amount. For the Truebit wallet, one green band of 8,535 ETH comes in and splits into red bands of 4,267, 4,001 and 268 ETH.

Trace Out adds the next step of the trail, and Trace In walks back toward the source. Each step saves a new version of the trace, so you can go back if a step gets messy.

The table under the chart shows Bitquery’s labels next to each address. Both wallets that took the big piles carry the tag hacker-truebit-hacker. Exchange wallets get tagged too. When I traced the Bybit hacker’s wallet from February 2025 the same way, the money came from a Bybit wallet tagged bybit-proof-of-reserves.

The chart also showed something I did not expect. A fan of tiny payments, each worth less than $1, came into the attacker’s wallet from look-alike addresses. One look-alike copies the first and last four characters of a wallet the attacker used, so both read 0x3b58鈥�862a in short form. The trick is called address poisoning. Scammers hope you copy the fake address from your history next time, and here they aimed it at the thief. A threshold box above the chart hides payments below any dollar amount you choose.

Two more buttons help when you need to share the work. Generate report turns the trace into a table of every transfer, with full hashes, which you can download as a CSV file. Get API opens the query behind the chart in Bitquery’s GraphQL IDE, so a developer can run it again.

Where it falls short. Solana is not in the chain menu yet. Each trace follows one currency, so if a thief swaps ETH for USDT, you start a second trace. And after a few steps the chart can turn into loops when money flows back into wallets already on it. In the Truebit case, the third step got hard to read, and the table was clearer.

2. Bitquery MCP: ask your AI assistant to follow the money

MCP, short for Model Context Protocol, is a standard way to plug outside tools into an AI assistant. Bitquery MCP plugs live blockchain data into Claude, ChatGPT, Cursor, VS Code and OpenAI Codex.

Setup is short. Add https://mcp.bitquery.io as a custom connector in your assistant, then sign in with your Bitquery account. There is no API key to paste.

You then get 314 tools. They trace transfers on 19 chains, including Solana, Bitcoin, Ethereum, TRON, the XRP Ledger and Stellar. They also cover DEX trades and prices on 10 chains. Guided prompts handle the common jobs: money_flow follows funds until they reach an exchange or a mixer, and risk_screen returns a LOW, MEDIUM or HIGH verdict with evidence.

I gave Claude the Truebit attacker’s wallet and asked where the ETH ended up. I also asked it to name every wallet on the way and score the last ones for risk. It pulled the transfers hop by hop, looked up the labels and ran a risk check on each final wallet.

The answer named two exit wallets. The first, 0x7720鈥59d, got 4,267 ETH after three more hops and sent 4,266.9 ETH to Tornado Cash on 10 January. The second, 0xd841鈥135, got the 4,001 ETH pile plus the 268 ETH payment and sent 4,268.5 ETH to Tornado Cash the same evening. Both scored 100 out of 100 for risk, the top of the scale, mainly because they paid a mixer. The risk check also flagged look-alike poisoning addresses next to both.

MoneyFlow reached the same Tornado Cash address three steps out from the attacker. The MCP got there from one question. Arkham has a hosted connector too, so the real difference is access. Arkham’s needs an approved API plan or trial, while Bitquery’s trial is self-serve and needs no card.

The product page listed three plans on 29 October 2026. The free trial runs 7 days with 100 credits, and it needs no card. Trading MCP costs $19 per month and covers trading data only, with 32 tools and no tracing. AI Investigation MCP, the plan with tracing, costs $149 per month, or $100 per month billed yearly, and includes all 314 tools.

Where it falls short. On Ethereum the risk check looks one step out from a wallet; on TRON it can look several steps. An AI assistant can also misread a result, so open the transaction hashes it cites before you rely on them. And like MoneyFlow, it follows one chain at a time, so you hop a bridge yourself.

Arkham vs Bitquery at a glance

In words: Arkham’s free plan shows the last 100 transactions per wallet, while MoneyFlow uses whatever date range you set and the MCP pages through the full history. Arkham’s Visualizer stays free, and MoneyFlow draws a flow chart too; the MCP answers in text.

Arkham Enterprise follows money across bridges for you. MoneyFlow and the MCP work one chain at a time. Risk scores come with Arkham Enterprise and with Bitquery MCP, not with MoneyFlow.

Price works differently on each side. MoneyFlow is free with a Bitquery account. Arkham Enterprise lists no price. Its API starts at $100 after approval, or $0.20 per credit pay as you go. Bitquery MCP is self-serve, with a free trial and then $149 per month for tracing.

5 more Arkham alternatives worth knowing

3. Chainalysis Reactor

Best for police, prosecutors and large exchanges. Reactor is Chainalysis’s case graph, and law enforcement uses it widely. There is no public price and no self-serve trial. If you work alone or on a small budget, it is likely out of reach.

4. TRM Labs

Best for compliance teams and government buyers. TRM sells its Forensics tool for tracing funds across chains, next to its wallet screening tools. Most of it is sold by quote, mainly to large firms.

5. Dune

Best for analysts who write SQL. Dune lets you query blockchain data and build dashboards, and it has a free tier and its own MCP server. There is no ready-made tracing screen, so you write the queries yourself or let its MCP server write them.

6. Etherscan

Best for checking one transaction or one wallet for free. Etherscan shows name tags on well known addresses, and it tags the Truebit attacker’s wallet “Truebit Exploiter 1”. There is no flow chart, so tracing a trail across many hops means a lot of tabs.

7. Nansen

Best for trading research. Nansen labels wallets and marks the ones with a strong record as “smart money”. You can then watch what they buy and sell. There is a free tier and a paid Pro plan, plus an API, an MCP server and a command-line tool. Its tools center on trading signals, so it is a weaker fit if you need to build a case file.

Other Bitquery products for investigators

Bitquery sells three more things for this kind of work:

• Coinpath: the money-flow data behind MoneyFlow, sold to companies with full history and bulk export.
• Crypto investigation service: if you were robbed and would rather hand the work to someone, Bitquery’s team traces the funds. Its page says the first assessment is free.
• Ethereum datasets: the full history of the chain as files, for teams that want to run their own analysis.

How to choose an Arkham alternative

• You want Arkham’s look without paying for Enterprise: Bitquery MoneyFlow, which is free.
• You already work in Claude or ChatGPT: Bitquery MCP.
• You trade and want smart-money signals: Nansen.
• You work for police or a regulated firm with a budget: Chainalysis Reactor or TRM Labs.
• You write SQL: Dune.
• You only need to check one transfer: Etherscan.

FAQ

Is there a free alternative to Arkham Intelligence?

Yes. Bitquery MoneyFlow is free with a Bitquery account and traces a wallet over any dates you pick. Etherscan is free for looking up one wallet or transaction at a time, and Dune has a free tier. Bitquery MCP has a 7-day free trial with 100 credits and no card. Arkham itself still has a free plan, but it shows only the last 100 transactions per wallet.

What is the best Arkham alternative for tracing stolen crypto?

For a visual trail, Bitquery MoneyFlow; for asking questions in plain English, Bitquery MCP. For court work with a large budget, Chainalysis Reactor or TRM Labs. I work at Bitquery, so weigh my pick with that in mind.

Does Arkham have an API?

Yes. You apply through Arkham’s API form, and once Arkham approves you, plans start at $100, billed in monthly credits. Some applicants get trial access first. Arkham also hosts an MCP server for AI assistants, and AI agents can pay per call at $0.20 per credit in USDC.

Can AI trace crypto transactions?

Yes, if the assistant is connected to live blockchain data. With Bitquery MCP, Claude followed the Truebit money to Tornado Cash and scored both exit wallets 100 out of 100 for risk. Check the transaction hashes it cites before you act on them.

Can you trace crypto after it goes into Tornado Cash?

Not directly. A mixer pools many people’s deposits, so the link between a deposit and a withdrawal breaks. Investigators match the timing and size of later withdrawals, which is slow and gives odds, not proof.

Try it yourself

Start a trace at lite.bitquery.io, or connect Bitquery MCP to your AI assistant and ask it to follow a wallet. If you want the raw history instead, Bitquery sells the Ethereum dataset as files.

Best Arkham Intelligence Alternatives in 2026: 7 Tools, 2 Tested on a $26M Hack was originally published in Coinmonks on Medium, where people are continuing the conversation by highlighting and responding to this story.