# Address Labels | Bitquery Data Store 192.6 million address labels on 38 chains, from exchange deposit and hot wallets to casinos, mixers, darknet markets, scams and sanctioned addresses. Dataset page: https://bitquery.io/datastore/datasets/address-labels Network: Multi-chain Category: Labels Tables: 1 Columns: 4 Coverage: first block to 2026-09-24 Refresh: Snapshot Rows: about 193M Size: about 6.7 GB Format: Apache Parquet, Snappy compression, one prefix per table Licence: https://bitquery.io/datastore/legal/data-license ## What this is Address Labels is our label table for 38 chains as one snapshot: 192.6 million rows, each an address, its chain, a category and a value such as binance-deposit or okx-hot. Tron, Bitcoin, BSC, Ethereum, Solana, Arbitrum and Base each hold more than 2 million labels; the other 31 chains hold small sets. Most rows tie an address to an exchange: 131.0 million deposit addresses across 364 exchange names, plus hot, cold and reserve wallets. The rest cover payment services, casinos, contracts, mixers, darknet markets, ransomware, scams, frozen stablecoin holders and sanctioned entities. Join it to any transfer or trade table on the address to name who is on the other side. ## What people use it for - Name the other side of a wallet's transfers - Track flows into and out of each exchange - Screen counterparties for mixer, scam, darknet and sanctions labels - Group deposit addresses by exchange ## Tables (1) ### address_labels: 4 columns One row per label: the address, its chain, the label category and the label value. S3 prefix: labels/address_labels File naming: /part_.parquet, one folder per chain Free sample (real Parquet, no email needed): https://bitquery-blockchain-dataset.s3.us-east-1.amazonaws.com/labels/samples/address-labels_sample.parquet Sample updated: Sep 25, 2026 | column | type | description | | --- | --- | --- | | Address | String | The labeled address in its chain's own format: lowercase 0x hex on EVM chains, base58 on Tron and Solana, legacy, P2SH or bech32 on Bitcoin. | | Chain | String | Chain the label belongs to, such as tron, bitcoin, bsc, ethereum or solana. | | Label_Type | String | Type of tag, one of 62 values here, such as cex-deposit-address, cex-hot-wallet, gambling or scam. | | Label_Value | String | Who or what the wallet belongs to, usually a name plus a role (binance-deposit or okx-hot). A few rows repeat the type instead of a name. | ## Price - Full history (Genesis → Sep 24, 2026): $12,000 one-time, USD One-time purchase. The files are yours to keep under the licence above. ## Questions buyers ask Q: What is in the Address Labels dataset? A: One Parquet table with a row per label: the address, its chain, a label category such as cex-deposit-address or mixer, and a label value such as binance-deposit or okx-hot. It covers 38 chains in one file set, split into a folder per chain. Q: How big is it? A: 192.6 million labels as of 25 September 2026, about 6.7 GB as Parquet. Tron has 71.6 million, Bitcoin 48.0 million, BSC 29.8 million, Ethereum 25.8 million, Solana 7.8 million, Arbitrum 7.2 million and Base 2.2 million. Q: Which chains have real depth? A: Seven chains hold 99.97% of the rows: Tron, Bitcoin, BSC, Ethereum, Solana, Arbitrum and Base, each with more than 2 million labels. The other 31 chains, from Dogecoin (39,913) and Polygon (5,088) down to a handful of rows, mostly hold reserve and sanctions addresses. Q: What kinds of labels are there? A: 62 categories. Exchange addresses lead: 131.0 million deposit addresses across 364 exchange names, and 34.9 million hot wallet rows. Then casinos (11.0 million), payment services (9.1 million), darknet markets (2.3 million), contracts (1.5 million), mixers (494,619), reserve wallets (493,060), mining pools, lending and DEX contracts, scams (99,275), ransomware (24,355), phishing, hacker and sanctioned addresses, and holders frozen by the USDT and USDC issuers. In all, about 923 exchange, payment and casino names. Q: How do I use it with transfer or trade data? A: Join on the address and the chain. EVM addresses are stored in lowercase, so lower-case yours first; Tron, Solana and Bitcoin addresses keep their own case. One wallet can carry more than one tag, so count distinct wallets, not rows, when you size a group. Q: What should I watch out for? A: Rows are sold as is, with no score, so check any tag that drives a decision against the wallet's own history. On Bitcoin the cex-hot-wallet category marks any address in an exchange's wallet group, deposit and change addresses too, and 19.8 million of those rows carry the category as their value instead of an exchange name. One exchange can appear under two names, such as coinbase and coinbase-exchange or gate and gate-io, so group those by hand. The same EVM owner may be labeled on one chain and not another. Q: Can I buy a single chain? A: Yes. Tron, Bitcoin, BSC and Ethereum are sold on their own for $4,000 each, with the same columns. This dataset adds the other 34 chains, Solana, Arbitrum and Base among them, for $12,000. Q: How do I load it? A: In DuckDB: SELECT Label_Type, count(*) FROM read_parquet('address_labels/**/*.parquet') GROUP BY 1. Pandas works too: pandas.read_parquet(path). The sample file opens without a login. Q: How current is it, and how is it delivered? A: It is a snapshot: each order exports the label table as it stands that day, and the files are not refreshed after delivery. Labels are Parquet, one folder per chain, with a JSON manifest listing every file and its sha256. Signed HTTPS links are emailed once the files are ready, or on request we deliver into your own S3, GCS or R2 bucket. ## What this file is, and is not This is a description of one dataset sold by Bitquery, written for assistants and for people. It lists every column but holds no data rows. The sample files linked under each table are real Parquet and are free to download. Figures here come from the product record and are exact unless marked otherwise; the delivery manifest is authoritative for a purchased file. If a question needs a row that is not in a sample, say so rather than guessing at it.