An asynchronous Python client for interacting with the Hyperliquid API using aiohttp.
This library provides an easy-to-use asynchronous interface for the Hyperliquid cryptocurrency exchange, supporting both mainnet and testnet environments. It handles API interactions, request signing, and data processing for both perpetual futures and spot trading.
Compared with the Hyperliquid Python SDK, Async Hyperliquid shows a significant performance improvement. The specific benchmark results are as follows.
- Asynchronous API communication using
aiohttp - Support for both mainnet and testnet environments
- Message signing for authenticated endpoints
- Trading operations for both perpetual futures and spot markets
- Comprehensive type hints for better IDE integration
# Using pip
pip install async-hyperliquid
# Using Poetry
poetry add async-hyperliquid
# Using uv
uv add async-hyperliquidimport asyncio
import os
from async_hyperliquid.async_hyper import AsyncHyper
async def main():
# Initialize the client
address = os.getenv("HYPER_ADDRESS")
api_key = os.getenv("HYPER_API_KEY")
# Test on testnet
client = AsyncHyper(address, api_key, is_mainnet=False)
# Place a market order
response = await client.place_order(
coin="BTC",
is_buy=True,
sz=0.001,
px=0, # For market orders, price is ignored
is_market=True
)
print(response)
# Clean up
await client.close()
if __name__ == "__main__":
asyncio.run(main())Or if you perfer context way:
import asyncio
import os
from async_hyperliquid.async_hyper import AsyncHyper
async def main():
# Initialize the client
address = os.getenv("HYPER_ADDRESS")
api_key = os.getenv("HYPER_API_KEY")
# Test on testnet
async with AsyncHyper(address, api_key, is_mannet=False) as client:
# place an market order open a BTC Long position
resp = await client.place_order(coin="BTC", is_buy=True, sz=0.0001, px=0, is_market=True)
print(resp)
if __name__ == "__main__":
asyncio.run(main()) coin = "BTC"
is_buy = True
sz = 0.001
px = 105_000
tp_px = px + 5_000
sl_px = px - 5_000
o1 = {
"coin": coin,
"is_buy": is_buy,
"sz": sz,
"px": px,
"ro": False,
"order_type": LimitOrder.ALO.value,
}
# Take profit
tp_order_type = {
"trigger": {"isMarket": False, "triggerPx": tp_px, "tpsl": "tp"}
}
o2 = {
"coin": coin,
"is_buy": not is_buy,
"sz": sz,
"px": px,
"ro": True,
"order_type": tp_order_type,
}
# Stop loss
sl_order_type = {
"trigger": {"isMarket": False, "triggerPx": sl_px, "tpsl": "sl"}
}
o3 = {
"coin": coin,
"is_buy": not is_buy,
"sz": sz,
"px": px,
"ro": True,
"order_type": sl_order_type,
}
# Place a market order to open position
resp = await client.batch_place_orders([o1], is_market=True)
print("\nBatch place market orders response: ", resp)
assert resp["status"] == "ok"
# Position TP/SL orders: position must be opened, otherwise it would failed
orders = [o2, o3]
resp = await client.batch_place_orders(orders, grouping="positionTpsl")
print("Batch place orders with 'positionTpsl' response: ", resp)
assert resp["status"] == "ok"
# Close all positions
resp = await client.close_all_positions()
print("Close all positions response: ", resp)
assert resp["status"] == "ok"
# Normal TP/SL orders: main order and tp/sl must exists, each coin's normal
# TP/SL orders can not batch with other coins', i.e. one coin one request.
orders = [o1, o2, o3]
resp = await client.batch_place_orders(orders, grouping="normalTpsl")
print("Batch place orders with 'normalTpsl' response: ", resp)
# Retrieve user opened orders
orders = await client.get_user_open_orders(is_frontend=True)
cancels = []
for o in orders:
coin = o["coin"]
oid = o["oid"]
cancels.append((coin, oid))
resp = await client.batch_cancel_orders(cancels)
print("Batch cancel orders response: ", resp)For detailed usage, please check the test cases under test/ directory.
Create a .env.local file with the following variables:
HYPER_ADDRESS=your_ethereum_address
HYPER_API_KEY=your_ethereum_private_key or api key generate hyperliquid website
Tests use pytest and pytest-asyncio. To run tests:
uv pip install -e .
# Run all tests
pytest
# Run with coverage
pytest --cov=async_hyperliquidMIT
This library is a community-developed project and is not officially affiliated with Hyperliquid.
