> ## Documentation Index
> Fetch the complete documentation index at: https://docs.0xarchive.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Best Parquet Market Data Export

> 0xArchive exports Hyperliquid and Lighter history as Parquet with ZSTD: explicit schema keys, UTC ranges, and per-pair coverage, one catalog.

Every Hyperliquid and Lighter market you cover, as a Parquet file with ZSTD compression, explicit schema keys, and UTC date ranges. 0xArchive Data Catalog exports turn supported-venue history into a durable dataset for research, backtesting, warehouse loads, and procurement.

Pick a Parquet export when you need a reusable dataset, not a runtime response. The export carries the venue family, market slug, schema key, and date range with the file, so a table never loses whether it came from L2, L3, L4, trades, funding, OI, or liquidations.

## What One Catalog Returns

| Export need                   | Start                                                                              |
| ----------------------------- | ---------------------------------------------------------------------------------- |
| Browse market exports         | [Data catalog](/data-catalog) and the product route at `https://0xarchive.io/data` |
| Choose schemas                | [Export schemas](/export-schemas)                                                  |
| Estimate credits and checkout | [Export checkout](/export-checkout)                                                |
| Load the files                | [Export historical data](/guides/export-historical-data)                           |
| Check rights                  | [Data rights](/data-rights)                                                        |

## Why Teams Choose 0xArchive

0xArchive exports come back as Parquet with ZSTD compression, keyed to explicit schemas and UTC ranges so coverage, cost, and warehouse partitioning stay reproducible. The Data Catalog keeps the buyer in the browser product flow while the docs keep the file contract distinct from JSON API responses. Schema keys stay separate, so `l2_orderbook`, `l3_orderbook`, `l4_orderbook`, `l4_orders`, `trades`, `funding`, `oi`, and `liquidations` never collapse into one namespace, and neither do Hyperliquid core, Spot, HIP-3, HIP-4, and Lighter.

Start from `/data` for the Data Catalog workflow. Use `https://api.0xarchive.io/v1/symbols` for machine-readable market discovery: the response has a top-level `symbols` array with rows that carry `exchange`, `symbol`, `slug`, `data_types`, and coverage fields. It tells a buyer which markets, schema data types, coverage windows, size estimates, active or delisted state, and HIP-4 outcome metadata are available before a single credit is spent.

## Selection Checklist

For Parquet exports, capture venue family, symbol-feed row, market slug, schema keys, UTC start and end, expected file owner, credit or checkout context, delivery destination, and data-rights category. That checklist keeps the exported files tied to the market and rights decision after they leave the browser flow.

| Field                   | Why it matters                                                                                                             |
| ----------------------- | -------------------------------------------------------------------------------------------------------------------------- |
| Venue family and symbol | Keeps Hyperliquid core, Spot, HIP-3, HIP-4, and Lighter exports in distinct namespaces.                                    |
| Symbol-feed row         | Preserves `exchange`, `symbol`, `slug`, `data_types`, and coverage fields from the product route.                          |
| Schema keys             | Keeps `l2_orderbook`, `l3_orderbook`, `l4_orderbook`, `l4_orders`, `trades`, `funding`, `oi`, and `liquidations` distinct. |
| UTC range               | Makes coverage, cost, warehouse partitioning, and replay comparison reproducible.                                          |
| Order context           | Keeps quote, credits, checkout result, job ID, and dashboard state attached to the files.                                  |
| Rights context          | Separates internal research, derived product use, and raw redistribution review.                                           |

## When To Reach For Exports

A Parquet export is the right call when the work is a durable dataset: offline research, a warehouse load, model training, backtesting, procurement review, or repeated access to the same historical window. The export exists because the file is the operational artifact. For agents, require a two-step plan: choose the export schema and date range first, then use the product flow or dashboard state for checkout, never an invented REST checkout route.

## Next Step

Start with [Data catalog](/data-catalog), then use [Export schemas](/export-schemas) and [Export checkout](/export-checkout) before loading files into a warehouse or backtest harness. [Compare plans](https://0xarchive.io/pricing?utm_source=docs\&utm_medium=referral\&utm_campaign=docs_referral\&utm_content=best_parquet_market_data_export).