mmflow-sdk 0.1.1

Python SDK for the mmflow market-data API — REST, SSE streams, and a typed WebSocket client.

mmflow — Python SDK

Typed Python client for the mmflow market-data API: REST endpoints, SSE streams, and a real-time WebSocket client. The REST + SSE surface is standard-library only; the WebSocket client needs the ws extra.

pip install mmflow-sdk          # REST + SSE (zero dependencies)
pip install "mmflow-sdk[ws]"    # adds MmflowSocket (websockets)

The package name is mmflow-sdk; the import name remains mmflow.

REST + candles

from mmflow import MmflowClient

mm = MmflowClient()

# OHLCV history (normalized to {time, open, high, low, close, volume})
for c in mm.candles("BTC", interval="1m", hours=24):
    print(c["time"], c["close"])

# Cross-venue open interest / funding / liquidations / whales
oi = mm.perps_oi(coins=["BTC", "ETH"], venues=["hl", "binance"])
for row in oi.data:
    print(row["coin"], row.get("oiUsd"))

# Anything else under /api/v1:
resp = mm.get("/quant/density", {"coin": "BTC"})

SSE streams

for ev in mm.stream_trades("BTC"):
    if ev.type == "trade":
        print(ev.data["price"], ev.data["size"])
    elif ev.type == "bye":
        break

Real-time WebSocket

import asyncio
from mmflow import MmflowSocket

async def main():
    sock = MmflowSocket(api_key="YOUR_KEY")
    sock.subscribe("trades", "BTC", lambda payload, meta: print(meta["channel"], payload))
    sock.subscribe("orderbook", "ETH", lambda payload, meta: print(payload))
    await sock.run()      # auto-reconnects; runs until close()

asyncio.run(main())

Channels: trades, orderbook, candles, liquidations, funding, openInterest, markPrice, footprint. Auth, exponential-backoff reconnect, and subscription replay are built in.