bsc_exchange_casino_flows
9 columnsTransfers between exchange hot wallets and casinos on BNB Chain, one row per exchange, casino and day.
bsc/exchange_casino_flows<YYYY-MM>.parquetone file per calendar month (ZSTD)Daily transfers between the hot wallets of 20 exchanges and 53 crypto casino brands on BNB Chain, one row per exchange, casino and day.
bsc_exchange_casino_flowsExchange to casino flows are hard to measure from raw transfers, because a player who withdraws from an exchange into a casino pays a one-off deposit address, not a known casino wallet. We find those deposit addresses each month, tie each deposit to a casino brand, and keep the deposits whose sender is an exchange hot wallet. One row is one exchange and one casino on one UTC day, with the number and USD value of transfers in each direction. It covers 1,084 hot wallets of 20 exchanges and 53 casino brands, both from Bitquery's own labels.
Most buyers sum deposit_usd by exchange for one casino, or by casino for one exchange. Only exchange hot wallets are matched. A casino payout into a player's exchange deposit address is not counted, so withdrawals are rare in this table, and money a player moves through a personal wallet between the exchange and the casino is not counted either.
Transfers between exchange hot wallets and casinos on BNB Chain, one row per exchange, casino and day.
bsc/exchange_casino_flows<YYYY-MM>.parquetone file per calendar month (ZSTD)Send this dataset’s full schema to an assistant and ask it anything. It reads the plain-text brief first, so the answer comes from the real column list rather than a guess.
Opens a new chat with the question filled in. Nothing about you is sent to Bitquery, and the assistant sees only the public brief.
Parquet files in a Requester Pays S3 bucket or a Cloudflare R2 bucket, chosen at checkout. On S3 you download with your own AWS account and AWS bills the transfer; on R2 you download with a read-only API key and pay no egress. Layout:
# bsc_exchange_casino_flows: one file per calendar month (ZSTD)
bsc/exchange_casino_flows<YYYY-MM>.parquetEvery window includes 9 columns. Only the time range changes.
bsc_exchange_casino_flows table, 9 columnsOne Parquet table with 9 columns: chain, date, exchange, casino, deposits, deposit USD, withdrawals, withdrawal USD and wallets. One row is one exchange and one casino on one UTC day. Tokens are BNB, USDT and USDC.
Small. A test run for Stake and Stake US on 10 November 2026 gave 8 rows: 7 exchanges paid into Stake and 1 into Stake US. We have not run a full month or all 53 casinos, so there is no month total to quote. Exact row counts and file sizes arrive with the delivery manifest.
20 exchanges: Binance, OKX, Bybit, Bitget, KuCoin, Gate, MEXC, HTX, Kraken, Coinbase, Crypto.com, BingX, CoinEx, BitMart, LBank, WhiteBIT, Bithumb, XT.com, CoinW and Bitkub, through 1,084 hot wallets in Bitquery's exchange labels. The casinos are the 53 brands of the BNB Chain Casino Player Wallets dataset, including Stake, BC.Game, 1xBet and Roobet.
Filter casino to the brand and sum deposit_usd by exchange. In the 10 November 2026 test, Binance hot wallets sent 484 deposits worth $85,758.59 to Stake, Bybit 99 worth $39,368.73 and MEXC 51 worth $15,560.30.
Only deposits sent straight from an exchange hot wallet. For Stake and Stake US on 10 November 2026 that was 687 of 2,447 deposits and $150,530 of $502,628. The rest came from other wallets, including players' own wallets that may have been funded from an exchange earlier.
A casino pays a player's exchange account through that player's exchange deposit address, and we match exchange hot wallets only. The withdrawals columns count casino transfers that go straight to an exchange hot wallet, which is uncommon. The test day had none.
We listed every counterparty of Stake and Stake US on 10 November 2026 with a separate query, matched them to the 1,084 exchange wallets offline, and got the same 8 rows as the dataset query, to the cent. That query ran with the 21 exchange wallets active that day; the form with all 1,084 wallets and all 53 casinos has not been run.
Deposit addresses are found per month, so a part-month finds fewer of them and its counts are a lower bound. Exchange labels are a snapshot, and a hot wallet an exchange started using later is missing until the list is refreshed. One hot wallet can serve several products of the same exchange.
With DuckDB: SELECT exchange, sum(deposit_usd) FROM read_parquet('bsc/exchange_casino_flows/*.parquet') WHERE casino = 'stake' GROUP BY 1 ORDER BY 2 DESC. pandas, Polars, Spark and ClickHouse read the same files.
A purchase covers the window up to yesterday, UTC, as flat Parquet files, one file per calendar month, with a JSON manifest listing each file and its sha256. We prepare the files within 7 business days of purchase and email you the location. You choose at checkout between a Requester Pays S3 bucket and a Cloudflare R2 bucket with a read-only API key. Files stay available for 7 days from delivery.
Yes. The table links to a public sample file with 8 real rows from 10 November 2026, for Stake and Stake US only.
192.6 million address labels on 38 chains, from exchange deposit and hot wallets to casinos, mixers, darknet markets, scams and sanctioned addresses.
Every bonding-curve trade, token creation, migration and pool on Pump.fun, plus per-token OHLCV, decoded and USD-priced.
Every TRX, TRC-10 and TRC-20 transfer on TRON since June 2018, one row per transfer, valued in USD where a price exists.
Tell us the chain, tables and date range.