Limits
We implemented several measures to ensure API stability during bursts of incoming traffic. If multiple requests are sent in quick succession, error responses may result as a status code 429.
To avoid overloading the API, it's important to observe these limits and handle 429 errors gracefully. If you notice an unexpected increase in rate-limited requests, reaching out to support is recommended.
We reserve the right to adjust limits either to deter misuse or to accommodate applications with high traffic. Should you need a higher rate limit, please contact support beforehand.
Connection limits
There is a limit of 1000 connections per attempt every 10 seconds.
The connection is per IP address.
How to interpret rate limits
Rate limits are accounted by specific intervals.
For instance, a 1 MINUTE interval begins at the start of each minute. A request submitted at 00:01:23.456 is counted towards the limit for the minute starting at 00:01:00. When the minute ticks over to 00:02:00, the count resets to zero.
This approach is consistent across different intervals. For example, a 1 DAY rate limit resets at 00:00 UTC daily, while a 10 SECOND interval resets at the start of every 10-second segment of a minute, such as at 00, 10, 20... seconds past each minute.
APIs may employ multiple rate-limiting intervals. If you deplete the quota for a shorter interval while a longer interval still permits requests, you must wait until the shorter interval's reset. Conversely, if you deplete a longer interval's quota, you must wait for its reset, regardless of the shorter interval's remaining allowance.
Exceeding the weight limit results in a 429 status code, signaling the need to halt further requests and avoid overwhelming the API.
Repeated violations of rate limits or failure to reduce request frequency after receiving 429 errors will lead to an automated IP ban, resulting in disconnection.
IP bans are monitored and escalate in duration for repeat offenders, ranging from 10 seconds to 24 hours.