Skip to main content
Every order, trade, and fill on Hyperliquid, live and historical, from one API. The native Hyperliquid API places orders and reads account state. For the market data, you call 0xArchive. One key returns order books, trades, candles, funding, open interest, liquidations, and full order-level (L4) depth as normalized JSON. Order books and trades go back to April 2023; order-level depth and lifecycle since March 2026. No collectors, no backfill, no rebuilding the book.

What One Key Returns

Hyperliquid market data0xArchive
Order books (L2), live and historical20-level snapshots about every 1.2 seconds since April 2023, 24.6B+ records, 176 symbols
TradesSince April 2023, continuous to the current hour
Funding and open interestSince May 2023
LiquidationsSince December 2025
Order-level depth (L4) and lifecycleEvery order add, fill, cancel, and TP/SL, with wallet attribution, since March 2026
Live stream and replayWebSocket subscriptions and exact-window replay on one socket
Spot, HIP-3, HIP-4Separate route families, each with its own history

Where The Native API Fits

0xArchive does not place orders; execution belongs to the venue’s own API, and for that job the native endpoint is the only correct choice. Order placement, cancels, and account state live on the exchange endpoint, signed by your wallet. The info endpoint also answers live market questions well: current mids for every coin, a 20-level l2Book snapshot, recent trades, and funding history, all free and venue-native. The bounds show up when the question is historical, and Hyperliquid’s own docs state them: candleSnapshot returns only the most recent 5,000 candles, userFills returns at most 2,000 recent fills, userFillsByTime reaches only the 10,000 most recent, time-range responses page at 500 elements, and l2Book is a live snapshot with no historical form. There is no request that returns the book as it stood last Tuesday. If the job is execution or reading your own account, 0xArchive is not the right fit; that stays on the venue. If the job is a backtest, a research window, or anything older than those retention caps, the native API cannot answer it.

When To Use 0xArchive

One call replaces a collector you would otherwise build and run against those retention caps. The market data comes back normalized, with named fields, live or historical: more than 24 billion order-book records and trades back to April 2023, plus full order-level (L4) depth with wallet attribution since March 2026. Every response carries a request ID and a data-quality path you can check before a backtest trusts the window. Access follows the key, not the plan: every route, market family, and schema is available on every tier, including Free; plans change credits, request rate, and replay speed, which scales from 10x to 300x by plan and up to 500x on Enterprise. Agents read the same surface through the hosted MCP server at https://mcp.0xarchive.io/mcp, a remote endpoint that takes your API key as a bearer token, nothing to install. Every order, trade, and fill. One API.

Run One Route

curl "https://api.0xarchive.io/v1/hyperliquid/orderbook/BTC/history?limit=1" \
  -H "X-API-Key: $OXARCHIVE_API_KEY"

Next Step

Open Hyperliquid REST, run one BTC request, then check the per-family windows on Venue coverage and Data quality before a job depends on one. Compare plans. For order-level depth, see Hyperliquid Order Book Data API.
Last modified on July 4, 2026