With a Grok model selected, the usage UI shows the Other models meter

Where does the bug appear (feature/product)?

Cursor IDE

Describe the Bug

With a Grok model selected, the usage UI shows the Other models meter. Grok 4.7, Grok 4.6, and Grok 4.5 are Cursor Models and should be shown against the Cursor Models pool.

Steps to Reproduce

  1. Sign in on a plan that includes both pools (Pro, Pro Plus, or Ultra).
  2. Open the usage UI (editor settings or the Spending tab).
  3. Select a Grok model (Grok 4.7, Grok 4.6, or Grok 4.5) and send a request.
  4. Watch which meter the UI presents for that usage.

Expected Behavior

Selecting Grok shows and updates the Cursor Models meter. The Other models meter stays unchanged for that usage.

Screenshots / Screen Recordings

Operating System

Windows 10/11

Version Information

Version: 3.22.12 (system setup)
VS Code Extension API: 1.128.0
Commit: 3a92974361033b2051526321308c2740fe5912c0
Date: 2026-09-26T04:55:12.125Z
Layout: Agent Window
Build Type: Stable
Release Track: Default
Electron: 42.10.0
Chromium: 148.0.7778.280
Node.js: 24.18.1
V8: 14.8.178.38-electron.0
xterm.js: 6.1.0-beta.291
OS: Windows_NT x64 10.0.26200

For AI issues: which model did you use?

Grok

Additional Information

Docs list Grok 4.7, Grok 4.6, Grok 4.5, and Composer 2.5 under Cursor Models, and third-party models under Other Models. Spillover onto Other Models is expected only after the Cursor Models allowance is used up. Subagent or helper calls that run a third-party model can also land in Other Models. This report is about the meter shown for the Grok request itself.

Does this stop you from using Cursor

No - Cursor works, but with this issue

Hey @lariozafilament, thanks for the report!

Your Grok usage is being counted correctly against the Cursor Models pool. The usage tray above the chat input is just showing the wrong meter when a Grok 4.6 or 4.7 model is selected. I’ve flagged this to the team. Sorry for the inconvenience here.

In the meantime, you can see the real numbers in Cursor Settings > Plan & Usage (the screen in your first screenshot). Your Grok requests go into the Cursor Models bar there, and Other Models only moves when you use third-party models like Claude.

Hey @lariozafilament,

The usage tray showing the Other Models meter when a Grok model is selected has now been fixed on our side, so no update is needed. Grok 4.6 and 4.7 usage should show against the Cursor Models meter.

Let me know if you still see the Other Models meter with Grok selected.

This same thing is happening for me as well. I use Grok 4.6 and 4.7 in 80% of cases. And yet I see that the ‘Other Models’ usage is being consumed right in front of my eyes.

This bug just started recently when my subscription was renewed for this month.

Hey @M_S4, thanks for the details.

I looked at your account, and your Grok 4.6 and Grok 4.7 requests are all being counted against the Cursor Models pool, so that part is working as expected.

What is moving your Other Models usage is Agent Review, the feature that reviews your changes for issues. It runs on a third-party model, so it draws from Other Models even when your chat is on Grok. Since your plan renewed, it accounts for all of your Other Models usage. It can start on its own after each commit, or after an agent finishes a turn that changed files, which is why the meter moves while you work.

If you’d rather it not run automatically:

  1. Open Cursor Settings.
  2. Go to the Git & PRs tab.
  3. Under Agent Review, turn off “Start Agent Review on Commit”.
  4. If you also see “Auto-Run on Agent Finish”, turn that off too.

You can still start a review manually whenever you want one. On your dashboard’s Spending tab these runs show up as agent_review, so you can check what they cost.

Let me know if Other Models keeps moving after that.

I found the Agent Review but it’s now called the ‘Source Control’ tab. I’m really surprised that you would not mention that this is charged like I would be using a third-party AI model, that’s a really important information . You should really add that important notice somewhere under the Agent Review tab. Is there any other setting left that I would be charged like I would be using a third-party AI model outside of Cursor?