[solved] Why is my GPT-5 mini 's rate 1x higher on Copilot Business ?

[hiro-gj]

Select Topic Area

Question

Copilot Feature Area

VS Code

Body

Hi, everyone.
Please help me..

Environment

• VS Code Version 1.109.0
  • GitHub Copilot Chat Version 0.37.1

Situation

• On February 9, 2026, when I checked the rates for each model in VS Code Copilot Chat,
  it was as shown in the image below.

*The number of available language models has also decreased...

• Since I have a "Copilot Business" subscription, the rate for GPT-5 mini should be 0.

Question

• How can I fix this?

• Below is the GitHub Copilot settings screen.
  Is this happening because any of these are set to "Enabled"?

• settings > copilot > features

  Details

  

• settings > copilot > policies

  Details

  

[Tamanna-Sharma8]

This is likely due to a pricing/policy update or a temporary policy sync issue in Copilot, rather than a local VS Code configuration problem.

A few things you can check:

1. Refresh Copilot authentication
Sign out of GitHub in VS Code and sign in again to force Copilot to refresh your license and policy data.

2. Confirm Copilot Business license assignment
Ask your organization admin to verify that your account is still assigned a Copilot Business license, as policy changes or license reassignment can affect model availability and rates.

3. Check organization Copilot policies
Some organizations restrict available models or change model access via policies, which may explain both the rate change and reduced model list.

4. Update extensions
Ensure both VS Code and Copilot/Copilot Chat extensions are fully updated, then reload the window.

If the rate still shows as 1× after policy and license verification, it may require GitHub support to refresh the license state on their side.

[hiro-gj]

Thank, you !

1. Refresh Copilot authentication
   Sign out of GitHub in VS Code and sign in again to force Copilot to refresh your license and policy data.

Oh... things have gotten worse.
However, it is clear that this is by no means Tamanna-Sharma8's fault!

[hiro-gj]

Hi! If you want, I can help troubleshoot further — is the issue with the rate still showing 1×, or something else in Copilot?

Thank you, Tamanna.

Currently, only Auto is displayed, but it is not able to reference any language models.
#186738 (reply in thread)

By the way, I found a similar issue,

Copilot Business (org seat): Premium models not visible in VS Code model selector after switching from Personal Pro #186439

Is this a widespread problem?

[hiro-gj]

This is a response/summary generated by Grok (xAI's AI).

Symptoms I'm seeing / others are reporting:

• Model selector shows almost nothing (only "Auto" or very few options)
• Grok-code / Grok Code Fast / other xAI models are completely gone
• Models like GPT-5 mini, Claude variants, etc. also vanished overnight for many people
• Happens on Copilot Individual, Business, and Enterprise accounts

What I've already tried (none fixed it for me yet):

• Ctrl+Shift+P → Developer: Reload Window
• Sign out of GitHub → fully quit VS Code → restart → sign in again
• Updated both GitHub Copilot + GitHub Copilot Chat extensions to latest
• Confirmed model settings at https://github.com/settings/copilot

Workarounds that helped some users recently (based on 2026 threads):

1. Fastest first try (most people recommend this first)
   Ctrl+Shift+P → run:
   Developer: Reload Window
   Wait 10–60 seconds and check the model picker again.

2. Full sign-out & clean re-login

   • Bottom-left status bar → GitHub icon → Sign out
   • Completely close VS Code (check task manager / activity monitor)
   • Reopen VS Code → sign in again

3. Delete Copilot cache folders (this fixed it for quite a few people)
   Close VS Code, then delete:

   Windows
   %AppData%\Code\User\globalStorage\github.copilot
%AppData%\Code\User\globalStorage\github.copilot-chat

   macOS
   ~/Library/Application Support/Code/User/globalStorage/github.copilot
~/Library/Application Support/Code/User/globalStorage/github.copilot-chat

   Linux
   ~/.config/Code/User/globalStorage/github.copilot
~/.config/Code/User/globalStorage/github.copilot-chat

   → Restart VS Code → sign in.
   (Backup the folders first if you're cautious)

4. Toggle models on GitHub settings page
   https://github.com/settings/copilot
   → Turn some models off → save
   → Turn them back on → save
   → Wait a few minutes and refresh VS Code

5. Just wait (unfortunately common right now)
   A lot of users said the models mysteriously reappeared after 6–72 hours with no action.
   Seems like a backend sync / rollout glitch on GitHub or xAI side.

Related threads (same symptoms reported):

• https://github.com/orgs/community/discussions/186738
• https://github.com/orgs/community/discussions/186439

[hiro-gj]

Hi, everyone, and Tamanna-Sharma8.
Thank you, I managed to solve it!

What I did

1. Update Windows to 25H2 and install VSCode from the MS Store.

   • Windows 11 Pro 25H2
   • VSCode 1.109.3
     • GitHub Copilot Chat 0.37.6

2. (This was also no good, so..) Let M365 Copilot Chat analyze my VSCode logs(*), which were initially enabled, then disabled, and finally enabled again.

[GitHub] (Organization) > settings > copilot > policies

• Copilot Chat in the IDE
• Copilot Agent Mode in IDE Chat

---

(*) M365 Copilot Chat analyze my VSCode logs.

(In fact, I ask questions and get answers in Japanese.)

I want to check for issues with GitHub Copilot chat, but which logs should I look at?

↓

First, check VS Code's Output.

Ctrl+Shift+P → Output: Show Output Channels
Select the following channels in order from the dropdown in the upper right corner to check their contents.

GitHub Copilot (extension log)
GitHub Copilot Chat (chat function log)
Log (Extension Host) (extension host log) as needed.

If necessary, increase Developer: Set Log Level to Trace to get more details (review the log again after reproducing the operation).
[Troubleshoot AI in Visual Studio Code](https://code.visualstudio.com/docs/copilot/troubleshooting)

↓

I wonder? Do you know why it's not working?

↓

Additional information on the current status (log timeline)

Token acquisition was successful at 09:09:46 (Got Copilot token). However, model enumeration failed due to chat not being enabled for IDE token.

An organizational policy warning (preview features are disabled by organizational policy) was also issued at the same time, so enabling the features for the organization is the top priority.