Hyperliquid

We run non-validating Hyperliquid nodes in both regions and read order statuses, fills and book state from the node, alongside the public WebSocket API.

Venue type

On-chain perpetuals exchange

Funding interval

Every hour, on the hour (interval_s = 3600)

Book checksum

Each L2 snapshot is checked against the book we rebuild from L3 events.

Sequence

L1 block height plus the event's index inside the block

Snapshots

A full-depth book snapshot at the top of every UTC hour (the 00:00 snapshot opens each daily file), plus the venue's own snapshots on reconnect, flagged with is_snapshot and snapshot_reason

Timestamps

Milliseconds (block time) in ts_exchange. ts_capture is int64 nanoseconds from a PTP-disciplined clock.

Regions

Tokyo (ap-northeast-1) primary, Ireland (eu-west-1) second

Participant id

User address (20-byte hex), as recorded on chain

Hyperliquid perpetuals are margined in USDC. Because the venue lists a single linear contract per asset, it resolves to BTC-PERP; BTC-USDC-PERP is accepted as an alias, and instruments.settle records USDC. The full rules are in symbology.

• Liquidations are flagged on the fill itself, with the liquidated address, so liquidations and book_l4 join on trade_id.
• Vault addresses, including HLP, carry entity_type = protocol in chain_entities, so with chain intelligence vault market-making flow in book_l4 can be separated from user flow.
• Funding, open interest and the oracle price come from the same block, so funding and mark_index share ts_exchange.

Extended fields are computed on ts_capture and published on a 100 ms grid. A field is available on a venue when its source data is.

The same calls work on every venue; only venue and symbol change. Endpoints go live with Phase 1, and the v1 spec is a draft until then.

from fdc import Client

fdc = Client(api_key="fdc_live_...")

# L4 order events with participant ids
book = fdc.book(
    venue="hyperliquid",
    symbol="BTC-PERP",
    depth="L4",
    start="2026-03-02",
    end="2026-03-02T01:00:00",
)
df = book.to_polars()

# funding at every settlement, then live
funding = fdc.funding(venue="hyperliquid", symbol="BTC-PERP",
                      start="2026-03-02", live=True)

curl -H "Authorization: Bearer fdc_live_..." \
  "https://api.financedatacloud.com/v1/book/l4?venue=hyperliquid&symbol=BTC-PERP&start=2026-03-02T00:00:00Z&end=2026-03-02T00:05:00Z"

{"op":"subscribe","channel":"book_l4","venue":"hyperliquid","symbol":"BTC-PERP"}

aws s3 cp \
  s3://fdc-archive/book_l4/venue=hyperliquid/date=2026-03-02/BTC-PERP.parquet .