How does Github Models Rate-Limiting work

[blazgocompany]

Select Topic Area

Question

Body

GPT 4o mini has a rate limit of 150/day
GPT 4o has a rate limit of 50/day
Does that mean there I can use 200 requests total over these two models?
Does that mean that all 25 models have their own rate limit, totaling to about 2000 requests/day?

[Sagaxious16]

Yes, you can use a total of 200 requests per day if you combine the rate limits of GPT-4o mini (150/day) and GPT-4o (50/day).

However, if each of the 25 models has its own rate limit, then theoretically, you could make up to 2000 requests per day (25 models * 80 requests/day). But this depends on the specific rate limits set for each model and whether they can be combined in this way.

[blazgocompany]

Thank you for clarifying that. Also is there a way to get the usage dynamically (from my Python script)? I know that it is possible to count each request sent, but is there a better way?

[talaqalotaibi]

Thanks for the explanation by asking the teams that use the public data. Models are calculated for them.