Query Hyperliquid funding rates with DuckDB SQL

The gold_funding_1h dataset (Pro) gives you hourly funding rates. Open it as a DuckDB relation and stay in SQL — DuckDB range-reads the Parquet and pushes predicates down to row groups.

import tessera

with tessera.TesseraClient() as client:
    rel = client.to_duckdb(
        "gold_funding_1h",
        coin=["BTC", "ETH"],
        month=tessera.MonthSpan("2026-01", "2026-05"),
    )

    # Average funding by coin and the most extreme hours.
    rel.aggregate("coin, avg(funding_rate) AS mean_funding", "coin").show()

Bring your own connection

Pass an existing connection to join Tessera data against your own tables:

import duckdb

con = duckdb.connect("research.db")
rel = client.to_duckdb("gold_funding_1h", "BTC", "2026-05", connection=con)
rel.create_view("btc_funding")
con.sql("SELECT * FROM btc_funding JOIN my_positions USING (time)").show()

Tip

Funding and positioning datasets are Pro-tier. A free key will raise tessera.ForbiddenError — upgrade here.