Ready-made datasets, derived from the full tape.
OHLCV candles, wallet PnL, liquidation history, and funding rates, computed from every fill and book delta we archive, not from samples. Subscribe to the dataset and skip the ETL pipeline you were about to build.
- OHLCV candles1s · 1m · 1h · 1d
- Wallet PnL63,000+ wallets
- Liquidation historysize · price · mark · victim
- Funding & basisper-market time series
All four computed from the complete tape: 1.65B+ order events archived, exact decimals throughout.
Candles built from every fill, not a sampled feed.
Per-market OHLCV at four resolutions, computed from the complete trade tape: 6.5M+ executed fills a day across 462+ markets. Exact decimal prices in, exact decimal candles out.
Because the candles and the live trade stream share one source, the last bar of your backtest and the first bar of production agree.
BTC perp · 1h candles · 48h sample window
hover for exact OHLCV
The leaderboard layer: PnL with names on it.
Realized PnL, traded volume, win-rate, and trade count per address, built from attributed fills, with every fee and rebate accounted for, across the 63,000+ wallets we’ve observed.
The raw tape carries the trader’s wallet on every fill; this dataset rolls it up. Copy-trading research, counterparty analysis, and flow studies start here instead of at a parser.
wallet-PnL dataset · sample rows · 30d window
click a header to sort
| 0x5dd8…9faf | +$1.8M | 57.7% | $373.1M | 84,171 |
| 0x1cc6…e6e7 | +$1.6M | 58.4% | $371.2M | 80,281 |
| 0x5001…e11b | +$1.4M | 70.0% | $208.4M | 37,961 |
| 0xcc9e…845c | +$584,847 | 57.6% | $229.7M | 53,889 |
| 0x8626…57c5 | +$560,779 | 67.4% | $84.8M | 73,173 |
| 0x4f86…6de2 | +$471,778 | 53.8% | $106.5M | 71,010 |
| 0xe69b…d3f8 | +$439,822 | 66.3% | $55.0M | 54,893 |
| 0x5eee…55bd | −$562,590 | 51.0% | $202.4M | 65,747 |
Sample rows in the wallet-PnL dataset format: 63,000+ wallets tracked across the full tape.
Every liquidation, with the victim attached.
One row per liquidation: size, price, mark price, and the liquidated wallet. Not an anonymized count: the full event, joined back to the fills it produced.
Risk models, cascade studies, and stress backtests need the quiet hours as much as the spikes. This dataset has both, continuously.
liquidation notional · hourly buckets · 72h sample window
hover for notional
Funding rates as a clean time series.
Funding-rate time series per market, sourced from the same ledger events the archive stores: rates and the payments they produced, exact decimals throughout.
Carry studies and basis trades live or die on the quality of the rate history. This one reconciles to the chain, for every market we cover.
BTC perp · funding rate · 72-point sample window
hover for exact rate
Built from the same exact tape.
Every dataset on this page is computed from the complete archive: every fill, every book delta, byte-exact decimals, full wallet attribution. Not sampled, not rounded, and it reconciles to the chain. When a candle looks wrong, you can open the raw fills underneath it.
Subscribe to the dataset, not the pipeline.
Take delivery through the query API, as bulk exports (Parquet / CSV / JSONL) to your bucket, or, on Enterprise, as scoped read-only database access.
Included with Ultra
All four derived datasets ship with the Ultra plan ($199/mo), alongside full historical query access and the monthly bulk-export quota.
Compare tiersÀ la carte subscriptions
Need one dataset, not a tier? Single-dataset subscriptions bill monthly per dataset, delivered via the query API or bulk export.
See à-la-carte pricingCustom datasets
A shape we don’t ship? Tell us the markets, date range, and schema; we build it from the tape and deliver it to your bucket. Quoted per request.
Book a DemoStart with the data, skip the backlog.
Self-serve keys cover the query API and Pro-tier datasets. For single-dataset subscriptions or a custom shape, talk to engineering.