Answer
The Trello REST API accepts a limit parameter whose maximum allowed value is 1000. Supplying a value greater than 1000 causes the API to ignore the supplied value and fall back to the default page size of 1000 items.
Confirmed fact
According to the official Trello API documentation, the limit argument is validated to be in the range 1‑1000 inclusive; any out‑of‑range value is treated as the default limit.
Likely explanation
When a client requests a page size larger than the allowed maximum, the server silently resets the limit to the default. This means that trying to “cheat” the rate limit by asking for a huge page does not reduce the number of round‑trips; the API will still return at most 1000 cards per request.
Optimizing the fields parameter
To reduce payload size and therefore the number of requests needed to pull a large board, request only the fields your application actually uses. The fields query parameter accepts a comma‑separated list of card attributes.
- Identify the minimal set of fields required (e.g.,
id,name,desc,due,labels,shortUrl).
- Include that list in the request:
GET /1/boards/{boardId}/cards?limit=1000&fields=id,name,desc,due,labels,shortUrl.
- If you only need identifiers, you can further reduce the payload to
fields=id,name.
- Combine field selection with a filter to limit the dataset, e.g.,
filter=open to exclude archived cards.
Putting it together – recommended request pattern
GET https://api.trello.com/1/boards/{boardId}/cards?
limit=1000&
fields=id,name,desc,due,labels,shortUrl&
filter=open&
key={yourKey}&token={yourToken}
This yields the maximum possible page size while keeping each response as small as practicable.
Mitigating 429 responses
- Honor the
Retry-After header (or X-RateLimit-Reset) returned with any 429 and wait that many seconds before retrying.
- Implement exponential back‑off if successive 429s occur.
- Monitor the number of requests per interval and ensure it stays below the documented limit for your API key/token (typically 100 requests per 10 seconds for a standard key).
Missing diagnostic detail
To fine‑tune the delay between requests, I need to know the exact rate‑limit quota applied to your API key/token (requests per interval). If this differs from the default 100/10 s, the recommended back‑off interval would change.