[FIX] Rate limits should be capped at 1000

[L-Agbaria]

Requests must have a maximum size of 1000, this is already something that is expected by the REST API for both runs and evaluations.

It should be updated for all OpenML objects even if the REST API currently handles it well, and this restrictions must be handled by the client to reduce the pressure on the server.

The documentation should also be updated to reflect this change, and the testing will need to go through verification after completion.

[satheesh142002]

Hi! I’m a first-time contributor and I’d like to work on this issue. Is it still available, or can I help with it?
Thanks!

[Jayant-kernel]

@L-Agbaria
Sorry to ask here but i couldnt find the community channel of the repo/org

[L-Agbaria]

@Jayant-kernel
This task covers datasets, tasks, flows, runs, evaluations, studies, and suites. The task itself is simple and hence the low effort, however this is something that must be done.

[Jayant-kernel]

Thanks for clarifying. I can take this on - I’ll cap request limits at 1000 across those objects, update the docs, and add tests.
Asking once again where is the community channel of the repo/org

[L-Agbaria]

I will also be participating more closely with this issue, as it is a multi-step goal.
The first goal is to first update all listing functions to use the shared list_all function, similar to what already exists for runs and evaluations. This includes (but is not limited to) looking up old issues and PRs related to this step, and possibly working on multiple files while keeping focus on the bigger picture.
I have increased the effort level to better reflect the scale of work that needs to be done despite the simplicity, if you are still interested, then let me know and we can organize a short meeting to discuss a plan next week. (This would also apply to you @satheesh142002)

[Jayant-kernel]

@L-Agbaria
I’m still interested. I’ll look through the related issues and PRs and map the listing functions that need to use list_all. A short meeting next week sounds good; let me know what time works for you.

[L-Agbaria]

Hello @Jayant-kernel
Would Tuesday 6 October 2026 14:00 (CET) work well for you?

Also @RugvedBane, would you be interested in joining?

[Jayant-kernel]

@L-Agbaria
Yes that works for me

[RugvedBane]

Hi @L-Agbaria, unfortunately I have college exams and a final year project right now so I won't be able to join the meeting. I would still love to contribute to this issue and if you have any small tasks feel free to hand them to me!

[Jayant-kernel]

@L-Agbaria
Small starting point for #1757. This caps the shared listing batch size at 1,000 and adds focused coverage for paged results.
Opening this as a draft so we can confirm the wider scope and any API-specific cases during the meeting before expanding the change.