Hyperliquid Info Endpoint Guide
The Info endpoint is Hyperliquid's read API for market, account, staking, portfolio, and historical state; correct integrations query the economic account, use the right asset identifiers, and respect endpoint-specific weights.
Hyperliquid's Info endpoint is the main REST surface for reading HyperCore data. It covers market metadata, mids, books, positions, open orders, fills, candles, funding, staking information, Portfolio Margin state, aligned quote status, and other protocol data. The challenge is not finding one universal response. It is choosing the correct query type and reconciling its scope with streamed state.
Market metadata first
Before submitting or interpreting orders, fetch metadata that defines assets, indices, size decimals, and market configuration. Perpetual and spot asset identifiers use different conventions in Exchange actions.
Hard-coded asset numbers are brittle when new markets are added.
Account queries
Use the master or subaccount economic address when reading balances and positions. An API-wallet address is a signer, not the account whose state it controls.
Account mode can also change which endpoint or state interpretation is relevant, particularly for Portfolio Margin.
Books and trades
The L2 book response exposes grouped price levels and can be configured by aggregation. Recent trade endpoints provide executions, not resting liquidity. Treating one as a substitute for the other produces incorrect liquidity models.
History limits
Endpoints can cap the number of returned records or candles. For research, paginate or persist data continuously rather than assuming one API call gives complete history.
Rate-limit weights also differ across Info request types, so a research crawler should schedule requests deliberately.
Build a typed data layer
Convert raw responses into internal typed structures with explicit units, timestamps, asset identifiers, and source names. Keep the transport layer separate from strategy logic.
That makes it easier to update the integration when a field or endpoint evolves and reduces unit-conversion mistakes.