Clarification Rate limited excceded Error

[Hunt092]

while developing an extension after a few calls i am randomly getting an rate limited exceeded error, in the documentation i see no clarification as to what the limit is, i am making sure that we don't cross the token limit per request, but is there a limit of how many request can be done by each account, is there any difference of the rate for business tier and enterprise tier.

How does the quota work, how will the request limit replenish, does the rate counter started with the first request? or is a time based (like x per hour )

[alefragnani]

Hi @Hunt092 ,

Could you provide more details about which service are you calling? Is it GitHub, ChatGPT or something else?

Thank you

[olguzzar]

tagging in @isidorn

[isidorn]

Thanks for opening this discussion.
I am curious what kind of extension are you building?

As for the tokens limit - we are constantly increasing those, and I think the new token limit size will be 64k.

As for the limit for the number of requests, we have a limit but let me find our exactly how large this limit is. Are you hitting this all the time? What are your reproducible steps?

[Hunt092]

Hi @isidorn,

Just as an overview of what we are doing, my team and i are building a transformation tool suite with Custom Chat participant so we go over multiple type so files, and have llm help us do some summarization / transformation, typically per file we have like 3x or 4x call, so once the process starts we will be making alot of calls to the extending of few hundreds for bigger chunks in a single command.

i am ensuring the token limit is not hit ( though having those extra tokens will be a blessing), but i am facing the rate limit issue when going for those bigger chunks of files,

i am also generally curious over how this rate limit works, is it time dependent, or there is a limit to parallel calls, is there a difference for enterprise vs business user for this limits

[isidorn]

There is not difference in limit for business vs enterprise that I know.
@thispaul can maybe point you to some documentation on GH Copilot limits.

[Hunt092]

HI @thispaul, can you provide any updates, or you if you need anymore information