Rate Limits

API rate limits and how to stay within them.

Rate limits are enforced per API key. Exceeding them returns 429 Too Many Requests.

---

Limits

Trading endpoints

Endpoint	Limit
POST /v1/orders (place order)	440 requests / 10 seconds
POST /v1/order/{id}/cancel	440 requests / 10 seconds
POST /v1/order/{id}/modify	440 requests / 10 seconds
POST /v1/orders/open/cancel (cancel all)	110 requests / 10 seconds
POST /v1/order/close-position	220 requests / 10 seconds
Global (all endpoints combined)	2,000 requests / 10 seconds

Query endpoints

Endpoint	Limit
GET /v1/orders/open	55 requests / 10 seconds
GET /v1/order/{id}	55 requests / 10 seconds

Public (unauthenticated)

Limit	Value
Max requests	20 per second per IP

---

When you’re rate limited

{ "status": 429, "message": "Too Many Requests" }

Stop immediately, wait at least 1 second, then retry with exponential backoff:

import time

def make_request_with_retry(fn, max_retries=3):
    for attempt in range(max_retries):
        response = fn()
        if response.status_code != 429:
            return response
        time.sleep(2 ** attempt)
    raise Exception("Max retries exceeded")

---

Use WebSocket instead of polling

The single most effective way to stay within limits is to stop polling and use WebSocket streams. One persistent connection replaces hundreds of repeated REST calls.

Don’t poll	Use instead
GET /v1/orders/open repeatedly	/v1/ws/private - SUBSCRIPTION_TYPE_ORDER
GET /v1/portfolio/positions repeatedly	/v1/ws/private - SUBSCRIPTION_TYPE_POSITION
GET /v1/account/balances repeatedly	/v1/ws/private - SUBSCRIPTION_TYPE_ACCOUNT_BALANCE
GET /v1/markets/{slug}/bbo repeatedly	/v1/ws/markets - SUBSCRIPTION_TYPE_MARKET_DATA_LITE
GET /v1/markets/{slug}/book repeatedly	/v1/ws/markets - SUBSCRIPTION_TYPE_MARKET_DATA

---

Cache reference data

Market and event metadata changes infrequently. Fetch it once on startup and refresh periodically rather than on every request.

import time

class MarketCache:
    def __init__(self, client, ttl=300):
        self._client = client
        self._cache = {}
        self._ttl = ttl
        self._last_refresh = None

    def get(self, slug):
        if self._needs_refresh():
            self._refresh()
        return self._cache.get(slug)

    def _needs_refresh(self):
        return not self._last_refresh or (time.time() - self._last_refresh) > self._ttl

    def _refresh(self):
        markets = self._client.markets.list({"limit": 100, "active": True})
        self._cache = {m["slug"]: m for m in markets["markets"]}
        self._last_refresh = time.time()

---

For automated systems

If you’re running an automated trading system and need higher limits for production:

1. Document your use case and expected request volume
2. Email [e-mail]
3. Include which endpoints you need higher limits for