Hyperliquid Historical Data API
Getting clean Hyperliquid history used to mean stitching raw S3 dumps, rate-limited venue calls, and a weekend of reconstruction code. 0xArchive launched with a single account and API key, plus routes that can be rerun. The initial archive covered Hyperliquid core perp history: order books, trades, funding, open interest, liquidations, and candles for builders who need the same window twice.
What you can pull
- Order books - point snapshots, historical windows, and L4 order-level routes.
- Trades - every fill with time, side, price, and size.
- Funding and open interest - derivatives context for backtests, dashboards, and risk views.
- Liquidations - liquidation events for risk and cascade analysis.
- Candles - OHLCV history for time-series work.
- Replay and exports - WebSocket replay and Data Catalog Parquet files for bounded jobs.
Data Coverage
Hyperliquid native order-book and trade history starts on April 15, 2023: more than 24.6 billion order-book snapshots and over 2.3 billion fills. Funding and open interest start on May 20, 2023.
Run one route
Create a key, then confirm the response shape before you wire anything around it:
curl "https://api.0xarchive.io/v1/hyperliquid/orderbook/BTC?depth=20" \
-H "X-API-Key: $OXARCHIVE_API_KEY"Then pick the surface that matches the job:
- Hyperliquid REST docs for route contracts and examples.
- Historical market data guide for bounded windows and request checks.
- Data Catalog for Parquet exports by market, schema, and UTC range.
- Pricing for plans, replay, and export credits.
What you get
The Hyperliquid backtesting workflow covers a fixed symbol and UTC window, cursor pagination, and an untouched holdout.
- REST API - route-family endpoints with cursor pagination for long windows.
- WebSocket replay - historical playback when event order matters.
- Python, TypeScript, and Rust SDKs - client libraries for REST and replay.
- Machine-readable surfaces - OpenAPI, llms.txt, MCP, CLI, and a Claude skill for generated clients and agents.