Skip to main content

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.

Trades and liquidations are separate event families. A venue can support trades without supporting liquidations in the same way, and a liquidation-volume summary is not the same dataset as raw liquidation events. Preserve venue family and route path in storage so downstream analysis does not mix unrelated sources.

Route Families

NeedStart with
Hyperliquid core trades/v1/hyperliquid/trades/{symbol}
Hyperliquid core liquidations/v1/hyperliquid/liquidations/{symbol} and /v1/hyperliquid/liquidations/{symbol}/volume
Hyperliquid Spot trades/v1/hyperliquid/spot/trades/{symbol}
HIP-3 trades/v1/hyperliquid/hip3/trades/{symbol} and /recent
HIP-3 liquidations/v1/hyperliquid/hip3/liquidations/{symbol} and /volume
HIP-4 trades/v1/hyperliquid/hip4/trades/{symbol} and /recent
Lighter trades/v1/lighter/trades/{symbol} and /recent

Usage Pattern

1

Confirm symbol family

A Spot pair, HIP-3 builder symbol, HIP-4 outcome side, and Lighter symbol need different routes.
2

Pick trade or liquidation data

Do not route a liquidation question to a trade endpoint just because both are historical events.
3

Check coverage

Trade and liquidation windows can have different coverage starts and access behavior.
4

Store request IDs

Backtests, heatmaps, and risk reports should preserve request IDs per page or window.

Exports And Examples

Data Catalog export schemas include trades and liquidations as separate choices. For parser behavior, use Responses. For bounded request examples, use Examples and the generated REST reference.

Parser Notes

Do not join trades and liquidations only by timestamp. Keep side, price, size, route family, source route, and request window with each row. For liquidation-volume summaries, store the aggregation route separately from raw liquidation events so reports can explain whether they used events, volume buckets, or both. Use recent-trade routes for short, current inspection and historical routes for bounded windows. Mixing those paths without a label makes backtests harder to explain later.

Raw Events Vs Volume

Use raw liquidation event routes, such as /v1/hyperliquid/liquidations/{symbol} or /v1/hyperliquid/hip3/liquidations/{symbol}, when the workload needs event-level liquidation records for audit, replay, or model input. These routes can be heavier because they return event rows and may need pagination, tighter windows, and a lower concurrency budget. Use /volume, such as /v1/hyperliquid/liquidations/{symbol}/volume?interval=1h or /v1/hyperliquid/hip3/liquidations/{symbol}/volume?interval=1h, when the workload needs aggregate liquidation-volume buckets for dashboards, scans, or backtests that do not inspect every raw event. For high-volume dashboards and repeated backtests, start with /volume and keep raw event pulls as a narrower drill-down path. For liquidation routes that return 429, honor Retry-After when present. If it is absent, use capped exponential backoff with jitter, reduce concurrency, and keep meta.request_id or x-request-id for the failed attempts. Do not retry unchanged after auth, access, malformed-request, unsupported-symbol, or invalid-parameter errors; fix the request or key before the next attempt.

Next Call After A Liquidation 429

When a raw liquidation route rate-limits, choose the next call based on the workflow instead of replaying the same request unchanged.
WorkflowSafer next call
Dashboard, scan, or repeated backtest summary/v1/hyperliquid/liquidations/{symbol}/volume?interval=1h or the matching HIP-3 /volume route
Event-level audit or model inputSame raw route with a smaller time window, lower concurrency, and request-ID logging
Unclear venue familyCheck Venue Coverage before switching between Hyperliquid core, Spot, HIP-3, HIP-4, and Lighter routes
Keep the failed raw-event path, Retry-After value when present, meta.request_id or x-request-id, and the chosen next-call route in the job log. That makes the retry decision reviewable when a backtest, alert, or export is handed to another engineer or coding agent.

Evaluation Fit

When a team says it needs “the tape,” confirm whether it needs raw trades, liquidation events, liquidation volume, or all three. A dashboard can often start with trades and summaries. A backtest, liquidation heatmap, or risk model needs route-level lineage and a data-quality check before results are saved.

Next Step

Open Examples for request patterns, Example Responses for payload shapes, Responses for envelope parsing, or Data Quality before using trades and liquidations in model inputs.
Last modified on May 19, 2026