Hyperliquid RPC
Hyperliquid runs a full exchange onchain, final in one block, across two layers that share one consensus and one state. HyperCore is the exchange layer: order books, matching, positions, and account state. HyperEVM is the smart contract layer: Ethereum-compatible, and able to read the books and trade on them directly. One Quicknode endpoint reaches both layers' RPC and streaming APIs. SQL Explorer uses a separate query endpoint and API key, and dedicated sentry peering for a node you run yourself is a flat monthly add-on.
What are you building?
Most jobs on Hyperliquid can be done more than one way. Trades arrive over gRPC, WebSocket, or as SQL history. Candles come from an endpoint or are built from fills. Each card lists the ways in.
Connect a non-validating Hyperliquid node you operate to three dedicated Quicknode sentry peers for blocks and full mempool data.
For HyperCore spot tokens and liquidity concepts, see the HyperCore overview.
Supported Hyperliquid Products
Deliver filtered HyperCore activity or HyperEVM blocks and logs to Postgres, S3, Kafka, or a webhook for downstream processing.
Subscribe to selected HyperEVM events and receive matching activity as HTTP notifications at your application endpoint.
Query 40+ indexed HyperCore tables with SQL to analyze trades, positions, funding, and ledger activity without running an indexer.
Connect your own non-validating Hyperliquid node to three dedicated sentry peers for raw blocks and full mempool data.
API surfaces
One Quicknode endpoint exposes every surface below. Start with the Quickstart, then use the HyperCore and HyperEVM overviews for endpoint routes, authentication, and examples.
Use /info to read market metadata, prices, exchange status, and user account state without sending a trade. It is the read-only entry point for market and portfolio views.
Build a trading action, sign it with your wallet, and submit it to HyperCore. The REST API also covers cancellations, preflight checks, and market discovery.
Use /hypercore to fetch historical HyperCore blocks, including raw replica_cmds, and batch account or order queries. It serves backfills and bulk portfolio reads.
Connect to /hypercore/ws for live trades, order updates, and book changes in a browser. Other streams cover TWAP, events, and writer actions.
Stream trades, order books, blocks, and mempool data over port 10000. Typed protobuf messages and zstd compression support low-latency, high-volume feeds.
Call contracts and read recent HyperEVM state through /evm. Use /nanoreth for historical queries, transaction traces, and debugging.
Which protocol serves which stream, and the gRPC versus WebSocket versus JSON-RPC decision, is in the available data streams reference.
For Python, TypeScript, Go, and Rust packages, see SDKs & Libraries.
Networks
/evm provides HTTP access to recent blocks and state. Older block data is pruned approximately every 12 hours; requests outside the retained range return invalid block height.
Use /nanoreth for historical queries, debug and trace methods, and HyperEVM WebSocket connections. Mainnet has full history; testnet history starts at block 34112653, not genesis.
Plans
HyperEVM and all 73 /info methods are on every plan, including the free trial. /hypercore JSON-RPC, WebSocket, and gRPC need a Build plan or higher. HyperCore method credits, stream limits, and plan access are on one page: Pricing & Plans.
Credit Usage
For RPC method credit costs, see the Hyperliquid API Credits table. That table does not include SQL Explorer queries or metered gRPC and WebSocket data. See Pricing & Plans for HyperCore gRPC and WebSocket streaming rates.
Rate Limits
Rate limits vary by plan tier. See the Hyperliquid rate limits and Quicknode plan pricing.
Specifications
Download the machine-readable API specifications for use with compatible tooling (e.g. OpenRPC Playground).