Quick Update on API Usage and Rate Limits

[Brian Maltzan]

HI all,

We want to share a brief update about the arXiv API service.

We are currently seeing very high API usage and increased load on our servers. Because of this, we need to delay any increases to current rate limits while we focus on improving server stability and reduce the 503 errors.

At some point, we are also planning a future replacement of the API platform to provide greater capacity and reliability. As part of that work, we will likely introduce API keys to help manage increased rate limits more effectively.

This arXiv-API Google Group is primarily provided as a community resource.  Our support team can be contacted here:

• arXiv Support Contacts

More information about our API policies and rate limits is available here:

• arXiv API Terms of Use
• arXiv API Documentation

Thank you for your patience and understanding.

Best regards,

Brian

[David W. Farmer]

Maybe you should charge for the high usage (which presumably are
from AI companies).

Regards,

David

On Thu, 14 May 2026, Brian Maltzan wrote:

>
> HI all,
>
> We want to share a brief update about the arXiv API service.
>
> We are currently seeing very high API usage and increased load on our servers. Because of this, we need to delay any increases to current rate limits while
> we focus on improving server stability and reduce the 503 errors.
>
> At some point, we are also planning a future replacement of the API platform to provide greater capacity and reliability. As part of that work, we will
> likely introduce API keys to help manage increased rate limits more effectively.
>
> This arXiv-API Google Group is primarily provided as a community resource.  Our support team can be contacted here:
>

> * arXiv Support Contacts

>
> More information about our API policies and rate limits is available here:
>

> * arXiv API Terms of Use
> * arXiv API Documentation

>
> Thank you for your patience and understanding.
>
> Best regards,
>
> Brian
>

> --
> You received this message because you are subscribed to the Google Groups "arXiv API Discussion" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to api+uns...@arxiv.org.
> To view this discussion visit https://groups.google.com/a/arxiv.org/d/msgid/api/CALi%2BNwqJ9VQ27vtmeqXcBv-BaFr0qGQmDeRgPGCisqd-Lt0gOg%40mail.gmail.com.
>
>

[Thorsten S]

arguably everybody everywhere is now using agents and arXiv should provide a MCP endpoint to its holdings

The model of operating locally synchronized copies of sub-sections or arXiv for specific purposes via OAI-PMH needs to be revised.

Best

Thorsten

To view this discussion visit https://groups.google.com/a/arxiv.org/d/msgid/api/9412b160-8b93-31fa-835-5905a4a4dc4%40aimath.org.

[David W. Farmer]

Dear Thorsten,

Yes, many people are using AI agents. That is not an argument
against charging each of those companies (each of which is
making money off the data they scrape) for access to that data.

Regards,

David

> To view this discussion visit https://groups.google.com/a/arxiv.org/d/msgid/api/CAMMiWBCyuoi1NfyEWSsGR_NCgQjfv2gmifZ%3DeF_DcwOuAYxZ%3Dg%40mail.gmail.com.
>
>

[Fabian Müller]

Hi Brian,

the third of your links is not working. I believe it should be

https://info.arxiv.org/help/api/index.html

(the "l" at the end was missing before).

Why do all three of these links have "?utm_source=chatgpt.com" attached at the end?

Many thanks for your superb work,
Fabian

On 5/14/26 21:11, Brian Maltzan wrote:
> HI all,
>
> We want to share a brief update about the arXiv API service.
>
> We are currently seeing very high API usage and increased load on our servers. Because of this, we need to delay any increases to current rate limits while we focus on improving server stability and reduce the 503 errors.
>
> At some point, we are also planning a future replacement of the API platform to provide greater capacity and reliability. As part of that work, we will likely introduce API keys to help manage increased rate limits more effectively.
>
> This arXiv-API Google Group is primarily provided as a community resource.  Our support team can be contacted here:
>

> * arXiv Support Contacts <https://info.arxiv.org/help/contact.html?utm_source=chatgpt.com>

>
> More information about our API policies and rate limits is available here:
>

> * arXiv API Terms of Use <https://info.arxiv.org/help/api/tou.html?utm_source=chatgpt.com>
> * arXiv API Documentation <https://info.arxiv.org/help/api/index.htm?utm_source=chatgpt.com>

>
> Thank you for your patience and understanding.
>
> Best regards,
>
> Brian
>

> --
> You received this message because you are subscribed to the Google Groups "arXiv API Discussion" group.

> To unsubscribe from this group and stop receiving emails from it, send an email to api+uns...@arxiv.org <mailto:api+uns...@arxiv.org>.
> To view this discussion visit https://groups.google.com/a/arxiv.org/d/msgid/api/CALi%2BNwqJ9VQ27vtmeqXcBv-BaFr0qGQmDeRgPGCisqd-Lt0gOg%40mail.gmail.com <https://groups.google.com/a/arxiv.org/d/msgid/api/CALi%2BNwqJ9VQ27vtmeqXcBv-BaFr0qGQmDeRgPGCisqd-Lt0gOg%40mail.gmail.com?utm_medium=email&utm_source=footer>.



--
FIZ Karlsruhe - Leibniz-Institut für Informationsinfrastruktur GmbH.
Sitz der Gesellschaft: Eggenstein-Leopoldshafen, Amtsgericht Mannheim HRB 101892.
Geschäftsführer: Andreas Schwartz.
Vorsitzende des Aufsichtsrats: MinR’in Dr. Elise Grauer.

[Brian Maltzan]

Hi Fabian, I was using chatgpt to check the grammar, it must have added that. Advertising? (I'm not a bot) Cheers,

> Why do all three of these links have "?utm_source=chatgpt.com" attached at the end?

To unsubscribe from this group and stop receiving emails from it, send an email to api+uns...@arxiv.org.
To view this discussion visit https://groups.google.com/a/arxiv.org/d/msgid/api/0bd9bfa2-499d-45ac-9409-15e3a264214c%40zbmath.org.