#
Rate limits
The API enforces rate limits to protect our servers against malicious and/or mistaken use. The API keeps track of the requests on an IP-by-IP basis. Hence, if you're on a VPN, proxy or a shared network in general, the requests of other users on this network might affect you.
At first, a global limit of 5 requests per second per IP address is in effect.
This limit is enforced across multiple load-balancers, and thus is not an exact value but rather a lower-bound that we guarantee. The exact value will be somewhere in the range
[5, 5*n]
(withn
being the number of load-balancers currently active). The exact value within this range will depend on the current traffic patterns we are experiencing.
On top of this, some endpoints are further restricted as follows:
Calling these endpoints will further provide details via the following headers about your remaining quotas:
#
Result Limit
Most of our listing endpoints will return a maximum number of total results that is currently capped at 10.000 items. Beyond that you will not receive any more items no matter how far you paginate and the results will become empty instead. This is for performance reasons and a limitation we will not lift.
Note that the limit is applied to a search query and list endpoints with or without any filters are search queries. If you need to retrieve more items, use filters to narrow down your search.
#
Further limitations
From time to time, we may apply undocumented limits to some endpoints if we deem the traffic patterns abusive. This can range from cache-busting attempts to persistent incorrect requests causing API error spikes.
Similarly, some specific request patterns associated with malicious intent result in instantly having your IP blocked for a period of time.
Finally, retrying requests upon being rate-limited will result in a combination of the rate limit being progressively extended until your requests are dropped entirely at the network layer.