Pro Users | What is the base model? I am confused

[brunopnogueira92]

Select Topic Area

Question

Copilot Feature Area

General

Body

I have something i need to be clarify because i couldnt understand yet.

So what is acctually the base model? Because i continually having a warnings saying that i reach a limit and i dont understand why. Right know the only information i have is that the base model is GPT4.1 but i still having the same warning.

Thanks in advance

[red-ca]

According to their docs, GPT-4.1 is the base model, are you using it with agent? because as far as i know agent is a premium request.

[gza]

I have a Pro+ subscription and see no "Base model" :

I supposed the GPT-4.1 is the base model, but it is accounted :

Where is the "Base Model" in the menu ? Does anyone sees it ?

[Virstriker]

There is no dropdown for downloading report and when i download full report it is not showing copilot.
@gza

[dajbear]

Click here

for the dropdown.

[Virstriker]

When i click it its showing directly this

[dajbear]

Different from mine. Try selecting this month then click email report.

[Virstriker]

Bro i have done that many times its not showing those stats that why i came here.

[rahulsamant37]

@brunopnogueira92 , Yeah and also you Can go through here - https://github.com/settings/copilot/features

[dajbear]

Github support you should join this conversation because this is very unclear. Even ChatGPT is questioning the legitimacy of this. How do we select the base model?

[gza]

The doc was updated :

link: https://docs.github.com/en/copilot/managing-copilot/monitoring-usage-and-entitlements/about-premium-requests

Good decision. It is not ready.

[mohammad-malik]

They should just scrap it.

[cuipengfei]

does this mean currently there is not limit? so pro and pro+ who paid diff gets same?

[Weckhawk]

Yes

[JavMB]

Hi! 👋

Great question - the model info changes frequently, so here's how to get the most current answer:

To check your current base model:

1. Open GitHub Copilot Chat
2. Ask: @github what model are you using right now?
3. It will tell you the exact model and version

Generally for Pro users you get access to the latest GPT-4 variants, but the specific version updates regularly.

You can also check the official GitHub Copilot docs for the most up-to-date model information, since these details change with each release.

Hope this helps you get the current info!

[gza]

We know model we use, it is not the point. We search for the "Base Model", the one which is supposed to not be accounted for Pro subscriptions.

Documentation looks clear. But when you look at the menu in VScode, it is impossible to know which one is the non-accounted "Base model".

[iwinux]

Let's be straight about it. The Pro plan offers "unlimited requests" of the base model. I want to use exactly that, unlimitedly. The question is: HOW?

[gza]

This is the question I asked to the support, I received only links to the docs as answer 😤

[geethasagarb]

Hi @brunopnogueira92 !

The “base model” in Copilot refers to the version of the AI model it uses — in this case, GPT-4. But when you see warnings about limits, it’s usually not about the model itself — it’s about usage limits tied to your Copilot plan.

Even with GPT-4.1, there are daily or monthly usage caps (especially for Pro users), and hitting those can trigger warnings. These limits help manage server load and fairness across users.

If you’re unsure what your specific limits are, check your billing or usage section in GitHub or Copilot settings. Hope this helps clear things up!

[juiicesb]

I understand that what you have is Copilot Free configured with GPT 4.1 in agent mode. You are probably getting the alert that you have reached the limit for code review requests, or whatever they call the option where the agent writes code for you. In Copilot Free, it’s 30 requests per hour, with restrictions on the number of lines, complexity and other internal limits wich i dont know...

This happened to me with Pro+ as well, when making many requests where the agent was making modifications for me. I reached a limit that blocked further requests for 2 hours.

[gza]

@juiicesb, @geethasagarb,

It is not a matter of rate limiting. It is a matter of accounting.

we search for the model corresponding to this line :

Because none of the ones in the menu matches. And GPT4.1 is in fact accounted (see screenshot)

[gza]

in this video www.youtube.com/watch?v=0Oz-WQi51aU at 35:31, we can see a "Copilot Base" model in the menu of the presenter:

Is it possible that a problem somewhere prevent this to be available for some users ?

Does anyone see this Copilot Base in his Agent model menu ?

[juiicesb]

github enterprise with copilot enterprise no option of copilot base, even in settings.

[gza]

Hi Everyone,

I got an answer from the Support ( thanks Eremie ) :

"The base model will ship once we ship premium request billing, so it’s not live yet. The reason why you're seeing "1x" for GPT-4.1 is because it is using the premium version and not the base model. The base model was supposed to be rolled out June 4th, but there have been some delays.

So let's wait until Github enables premium request accounting. Then we can check if there is such "Base Model".