What One Key Returns
| Market-data job | 0xArchive |
|---|---|
| Hyperliquid order books (L2), live and historical | 20-level snapshots about every 1.2 seconds since April 2023, 24.6B+ records, 176 symbols |
| Order-level depth (L4) and lifecycle | Every order add, fill, cancel, and TP/SL, with wallet attribution, since March 2026 |
| Hyperliquid families | Core perps, Spot, HIP-3, and HIP-4, each its own route family with its own history |
| Lighter | Trades, funding, and open interest since August 2025; L3 order-level depth since March 2026 |
| Trades, funding, open interest, liquidations | Hyperliquid since April 2023 (liquidations since December 2025), Lighter since August 2025 |
| Live stream and replay | WebSocket subscriptions and exact-window replay on one socket |
| Coverage, freshness, incidents, latency, gaps | One data-quality surface across every supported family |
Where Hydromancer Fits
0xArchive sells bounded market-data routes, not builder-platform operations; that operational layer is Hydromancer’s. Its site lists a builder REST API with batch endpoints such asbuilderFillsByTime and builderLiquidations that query 1,000+ wallets at once, WebSocket streaming for fills and orders, L2 and L4 order-book streaming with address visibility, and Reservoir, a free Parquet-on-S3 archive of Hyperliquid historical data. Paid tiers run 2,500 a month with an uptime SLA, and more than 80 apps build on it. When the job is builder-platform tooling, builder-fill analytics, or a free S3 corpus to explore, Hydromancer owns that job, and 0xArchive is not the right fit for it; this API is the archive layer underneath, the records rather than the platform.
Hydromancer covers one venue. When the same job also needs Lighter, or needs a bounded REST route with a stated historical window instead of a bucket to crawl, the decision moves here.