I am currently using different OpenAI models: gpt-4, gpt-4o, gpt-4o-mini, gpt-3.5-turbo, and embedding models. There is a sudden increase in usage shown under the ‘Other Models’ category in the usage section, and the price is also higher since August 16th.
For the other models, I have separate graphs and sections to show their usage and price, but for the ‘Other Models’ category, the usage is quite high.
We are not able to identify which model’s usage is high under ‘Other Models’ category. How do we identify what categories/models come under ‘Other Models’ and how they are being consumed?
Is anyone else having the same problem? I am being overcharged for models that I do not use.
It looks like there are others experiencing your issue.
I sent a message to OAI to look into this.
Yes, @AaryaJake
I faced the same issue recently. The sudden increase in usage under the ‘Other Models’ category and the associated cost is puzzling. I also have separate tracking for different OpenAI models, but this ‘Other Models’ category seems to aggregate usage in a way that makes it hard to pinpoint which specific model is responsible.
I definitely ran into the same issue.
I barely use the APIs and I only use 4o-mini. All of a sudden my credits dried up and was automatically topped up.
When I checked, there’s this huge spike in “other models” category that no one on my team is able to identify quite what it is.
Contacted OAI team, still waiting for a response.
Did anyone get to the bottom of this? We’ve had significant “other modeling” billing showing up this past week – and can’t identify what this is for.
Hey same too only this week…
What’s the reply? Problem is this issue reappeared even after deletng all users and changing all api keys. And it goes way over the limit with no problem…there is no safeguard form this…
Hi,
You can get help from the support system at help.openai.com, they will get back to you. The Developer forum is run by volunteers with no access to accounts or billing.
Also, if you can, I could suggest removing auto-recharge as a temporary measure if you have concerns over further charges.
There are people with alike issues…we need open AI to actually reply.
I am also facing the same issue.
Is there any way to tackle the problem without limiting the monthly budget? I am kind of middle of something and my monthly budget is depleting very fast
Have you tried checking the detailed activity at https://platform.openai.com/organization/usage/activity ?
First check/select the models, then export a CSV to see the detailed model names and usage.
I have same issue starting from 10th November 2024, trying to get help from the support system at help.openai.com; Still waiting for a response from Openai.
From the messages that arrive on the forum support can take a few days to respond.
I am noticing that my last message was exactly 3 weeks ago so it being a weekend when you messaged may have a baring.