Copy Trading Hyperliquid Top Traders: A Python Bot Tutorial

Why Follow Other Traders Instead of Building Your Own Strategy?

Copy trading Hyperliquid top traders is the fastest shortcut to exposure against proven alpha. Instead of back-testing signals for months, you let the market surface traders who are already profitable on-chain — every entry, exit, and PnL is publicly visible on Hyperliquid — and mirror their moves with a lag-free API feed.

The challenge isn't copying. The challenge is which trader to copy. The leaderboard is dominated by high-leverage gamblers whose one lucky month looks like genius until the drawdown. Real alpha requires filtering by win rate, profit factor, trade count, and duration — and those filters need data most leaderboards don't expose.

This post walks through: how we score Hyperliquid traders, how to build a copy-trading bot that only mirrors the filtered set, and the risk controls that matter. Endpoints: /api/v1/hyperliquid/top-traders and /api/v1/hyperliquid/copy-signals. Pro Plus tier.

How We Score Hyperliquid Traders

Each tracked address gets a composite score (0-100) based on 30-day performance. The components:

• Win rate — fraction of closed trades profitable. Useful but deceptive alone (a scalper can win 90% and still be unprofitable).
• Profit factor — gross winners / gross losers. Anything > 1.5 is credible; >3.0 is exceptional for 30-day sample.
• Trade count — 50+ trades over 30 days is enough to trust; <10 is statistical noise.
• Average duration — filters out HFT/scalpers you can't realistically copy given tx latency.
• Trader type — classified as hft, scalper, day_trader, swing_trader, or position_trader.

The scoring hits /api/v1/hyperliquid/top-traders. You filter it down to a watchlist via /api/v1/hyperliquid/watchlist/auto, which auto-refreshes daily so your list stays current as traders fall off or emerge.

How Do You Build a Copy-Trading Bot?

Three building blocks. Each is a single API call, plus a router and an executor.

1. Pick traders to follow. Call /api/v1/hyperliquid/top-traders, filter by score > 85 and trader_type != "hft", take the top 10.
2. Poll for their signals. Call /api/v1/hyperliquid/copy-signals every 60 seconds. Each response row is a structured {type: 'ENTRY'|'EXIT'|'SIZE_CHANGE', coin, side, size, entry_price, leverage}.
3. Route the signal to your execution venue. Could be Hyperliquid itself, Binance perps, or an on-chain DEX. Scale size down by your risk budget (typically 1-5% of the followed trader's notional).

The copy-signals endpoint only returns signals for traders who have recent activity in the last N minutes (default 60). You never see stale noise, only fresh moves.

Copy Trading vs Auto-Following Signals: The Key Difference

“Copy trading” and “signal auto-following” sound identical but have meaningfully different risk profiles.

On-chain copy trading on Hyperliquid is strictly better for any systematic workflow.

Python Bot That Mirrors Filtered Top Traders

A minimal copy-trading loop. This version only logs trades — wire your execution venue into route_signal() when you're ready.

import time
import httpx

API = "https://cryptodataapi.com"
HEADERS = {"X-API-Key": "cdk_live_YOUR_PRO_PLUS_KEY"}
SEEN: set = set()

def route_signal(trader_addr, sig):
    # Plug in your execution logic here
    print(f"{sig['type']:12s} {sig['coin']:6s} {sig['side']} sz={sig['size']} "
          f"@ {sig['entry_price']} lev={sig.get('leverage')} trader={trader_addr[:8]}")

while True:
    r = httpx.get(f"{API}/api/v1/hyperliquid/copy-signals",
                  params={"min_score": 85, "minutes": 2},
                  headers=HEADERS, timeout=15)
    for trader in r.json().get("traders", []):
        for sig in trader["signals"]:
            key = (trader["address"], sig["type"], sig["coin"], sig["size"], sig["entry_price"])
            if key in SEEN:
                continue
            SEEN.add(key)
            route_signal(trader["address"], sig)
    time.sleep(60)

The endpoint returns only traders with activity in the last minutes window, so the bot never wastes cycles on idle accounts.

Risks, Filters, and Controls

Copy trading isn't free money — it's inherited edge with inherited risk. Controls that matter:

• Trader-level risk budget — never risk more than 1-2% of your account per copied trader. Diversify across 5-10 traders to kill idiosyncratic blow-ups.
• Hard exclude HFT — set exclude_types=["hft", "scalper"] on the watchlist endpoint. You cannot beat their latency.
• Max-leverage cap — if a trader goes 20x on a position, your copy should still cap at 3-5x. You don't have their conviction or information.
• Drawdown circuit breaker — if your copy account loses more than 5% in a day, pause routing until you review. Auto-failure is the main operational risk.
• Auto-refresh the watchlist — set auto_refresh: true. Traders cool off quickly; yesterday's top-scorer can be tomorrow's also-ran.

Pair with our Hyperliquid perps data guide for full market context. For the signal-discovery side of the same workflow, see whale tracking and smart money flows.