Export Hyperliquid and Lighter Market Data

Yes. A Data Catalog export is a one-time file order priced by quote; you do not need an API subscription. Sign in to review and pay, then track the export in your dashboard.

Configure a Parquet export

Open the Data Catalog, select the exact venue deployment and market, choose eligible schemas and set UTC dates. HIP-4 also requires the event and Yes/No side selection. Review the cart so the same symbol on two venues does not become an accidental substitution.

Configure a Parquet export

Each schema lists its first exportable date. With multiple schemas or outcome sides, the shared range begins at the latest selected start; settlement can cap the end. The final quote enforces export-policy floors, which can be later than retained API history. An active market ending today is not a promise that today's data is finalized.

Configure a Parquet export

The quote lists size, rows and price per market/schema line. Price uses uncompressed GB, the schema's rate and minimum, with a $10 order minimum. Export credits apply first, then volume discounts on the remaining subtotal. Review the final total rather than treating the catalog's daily-size estimate as a fixed price.

Configure a Parquet export

Yes. Build includes $50, Pro $300 and Scale $1,500 in monthly export credits. They are a dollar balance, separate from API credits, and apply before checkout. Monthly allocations roll over without expiry up to twice the monthly allowance.

Configure a Parquet export

Apache Parquet with ZSTD compression, with a README describing the exact columns and types. Books can include companion checkpoint files: Hyperliquid L4 book exports include checkpoints, and Lighter L2 includes full-book snapshots. Preserve the package, not just the change file.

Configure a Parquet export

Read Parquet with pandas.read_parquet in Python or read_parquet in DuckDB, and use your warehouse's Parquet loader for bulk ingestion. Check the file-column reference for decimal, timestamp and side fields before joining. Keep exchange metadata because files for the same symbol and schema on two venues can share a filename.

Configure a Parquet export

Open dashboard/exports and select the order. It moves from Awaiting Payment to Queued, Processing and Ready to Download; an email also announces ready files. The order page lists Parquet files and README downloads. Download availability lasts 30 days after completion; signed links last 7 days and refresh when you reopen the order.

Configure a Parquet export

Queued and Processing mean the export is still being built. A paid failure shows Unavailable: export credits return automatically within a few minutes, while a card-paid failure needs support with the order ID for a refund. An unpaid expired order also shows Unavailable but has no payment to refund; rebuild it in the catalog.

Configure a Parquet export

Yes. Reopen the market configuration, select the new UTC range and review a new quote and order. A previous download is a fixed selection, not an automatically updated dataset. For scheduled file delivery, discuss Enterprise delivery rather than expecting one order to refresh itself.

Configure a Parquet export

Select the retained market in the catalog and check each schema's window. A settled HIP-4 event ends at its settlement date; a delisted market can still have retained exportable history. Settlement or delisting does not guarantee every schema exists.

Configure a Parquet export

The commercial Data License permits internal research, trading, storage and model training, plus materially transformed products that do not expose or reconstruct source records. Benchmarks, indices, financial products, standalone/substitute data feeds, raw redistribution and white-label/OEM data delivery require written permission. Downloaded entitled data can be retained and used after cancellation under the same license.

Read the Data License

API access pays for ongoing REST and WebSocket usage under your account plan. A Data Catalog order buys the selected historical Parquet files at a final quote. Pick the API when your application needs repeated reads, or the catalog when your next step is loading a bounded dataset into a notebook or warehouse.

Configure a Parquet export