defipipe 0.4.1

Historical DeFi data API client: per-block, point-in-time contract state as pandas DataFrames. Aave rates, Uniswap prices, USDC supply, aligned onto one time axis.

defipipe: historical DeFi data API, per block, as pandas DataFrames

What read functions actually returned at every Ethereum block: Aave borrow rates, Uniswap slot0 prices, USDC total supply. Reconstructed historically, indexed live, and aligned onto one clean time axis.

Most DeFi data is built from events. defipipe serves state: the result of calling a contract's read functions at a specific block. If you need to know what the Aave WETH borrow rate actually was at block 19,000,000, not an hourly average from a dashboard, this is the API for it.

Install

pip install defipipe

Five lines to an aligned DataFrame

from defipipe import Client

df = Client().data([
    "eth:aave:v3:0x87870bca3f3fd6335c3f4ce8392d69350b4fa4e2/weth:call:getreservedata[currentvariableborrowrate]",
    "eth:lido:v2:0xae7ab96520de3a18e5e111b5eaab095312d7fe84:call:gettotalpooledether",
], freq="1h", since="2026-07-07", until="2026-07-12")

df.corr()
df.attrs["units"]   # scaling recipe per series: unit, decimals, kind
df.attrs["blocks"]  # per-point source block per series

No API key needed for recent data at hourly bars. Full history and per-block granularity: get a key at https://defipipe.io/pricing and set DEFIPIPE_API_KEY (or pass api_key=).

Why quants use this

• Point-in-time correctness. Every value is the last observation at or before each interval boundary, resolved through the block↔timestamp map. No interpolation, no lookahead bias: safe for backtests.
• Exact raw values on demand. data(..., raw=True) returns the exact on-chain integers as Python ints (arbitrary precision, uint256-safe); the default float path requests lean payloads and skips them. Scaling is metadata you can audit, not arithmetic you have to trust.
• One key for everything. Canonical series IDs ({chain}:{protocol}:{version}:{instance}[/{scope}]:{kind}:{metric}) name every series across protocols; browse them at https://defipipe.io/datasets.
• Per-block granularity back to each pool's deployment, from our own archive node, not sampled snapshots.

Per-block rows, exact integers

raw = Client(api_key="dp_...").data(
    ["eth:uniswap:v3:0x88e6a0c2ddd26feeb64f039a2c41296fcb3f5640:call:slot0[sqrtpricex96]"],
    freq="block", since="2026-07-11", until="2026-07-12", raw=True,
)
raw.head()  # one row per block; values are exact Python ints

Windows of any size work: the client pages through the API cursor transparently.

Coverage

Aave V3, Uniswap V3, Lido today; Morpho, GMX V2, Uniswap V4, Ethena, Maple next. Request a pool at https://defipipe.io/feature-requests; approved requests ship with full backfilled history.