Agent Review Feature Missing Model Visibility and Selection

Ideas Feature Requests
1

Where does the bug appear (feature/product)?

Cursor IDE

Describe the Bug

The Agent Review Feature does not display which AI model is being used for code reviews, making it impossible to verify model selection or control costs. This creates a risk of unintentionally using expensive models for reviews.

Steps to Reproduce

Go to the right-side pane in agent mode, select the VCS tab up top and look for the agent review section, then hit review

Expected Behavior

Expected Behavior

  • The review pane should clearly display which model is currently being used
  • The review pane should have its own model selector, independent of the main Agent pane
  • Users should be able to select an appropriate (cost-effective) model specifically for reviews

Current Behavior

  • The Agent Review pane does not show which model is being used for reviews
  • It’s unclear if reviews use:
    • Composer 1
    • Another default model
    • The model selected in the main Agent pane
  • No separate model selector exists for the review pane

Operating System

MacOS

Current Cursor Version (Menu → About Cursor → Copy)

Version: 2.0.77
VSCode Version: 1.99.3
Commit: ba90f2f88e4911312761abab9492c42442117cf0
Date: 2025-11-13T23:10:43.113Z
Electron: 37.7.0
Chromium: 138.0.7204.251
Node.js: 22.20.0
V8: 13.8.258.32-electron.0
OS: Darwin arm64 24.4.0

Additional Information

Impact

Business Impact

  • Cost Risk: Potential for unintentional use of expensive models (e.g., Claude Opus, GPT-4) for reviews
  • Lack of Transparency: Cannot verify which model is processing reviews
  • No Cost Control: Unable to optimize review costs by selecting cheaper models

User Impact

  • Uncertainty about review quality vs. cost trade-offs
  • Inability to make informed decisions about model selection for reviews
  • Potential for unexpected billing surprises

Use Case

As a developer using Cursor for code reviews:

  1. I want to see which model is being used for Agent Reviews
  2. I want to select a cost-effective model specifically for reviews (separate from my main coding model)
  3. I want to verify that reviews are using my selected model to control costs

Proposed Solution

Option 1: Separate Model Selector (Recommended)

  • Add a dedicated model selector dropdown in the Agent Review pane
  • Display the currently selected model prominently
  • Allow independent model selection for reviews vs. main Agent pane
  • Show model cost/pricing information in the selector

Option 2: Model Indicator (Minimum)

  • At minimum, display which model is currently being used in the review pane
  • Show model name/identifier clearly visible during reviews

Additional Context

  • This is a critical feature for teams managing AI tooling costs
  • The review pane appears to be a separate feature from the main Agent pane, so it should have independent model selection
  • Model visibility is essential for transparency and cost management

Does this stop you from using Cursor

No - Cursor works, but with this issue

2 Likes
2

Hey, thanks for the report. You’re right that Agent Review currently doesn’t show which model is being used and doesn’t let you choose a separate model for reviews.

Right now, Agent Review runs as a normal agent operation and is billed as a usage‑based request. The model used usually follows your current Agent settings, but I understand your concerns about transparency and cost control.

This is a feature request rather than a bug: the system is working as designed, and you’re asking for better visibility and the ability to choose a separate model for reviews. These are very reasonable requests, especially for teams that closely manage AI spend.

A couple of clarifying questions that might help:

  • Have you checked your usage dashboard to see the actual cost impact of Agent Review operations?
  • Are you currently using “Auto” mode or a specific model in the main Agent panel?

Your detailed writeup with use cases and suggestions is very helpful. The team monitors these feature requests, though I can’t promise a specific implementation timeline.

3

Great follow up questions, and sorry for mislabeling as a bug, this is definitely an enhancement/feature request.

Have you checked your usage dashboard to see the actual cost impact of Agent Review operations?

The usage and analytics dashboard is all wrong. I don’t think it counted any agent reviews if they were triggered by the right pane. At least that’s where I’d start to look if trying to look for loopholes in the analytic tracking. For example, I must have kicked off at least 10 reviews yesterday or more but only 1 is showing from 1 user in the analytics dashboard.

Are you currently using “Auto” mode or a specific model in the main Agent panel?

No, I’m selecting either sonnet 4.5, gemini 3 pro or composer 1. Sometimes in max sometimes not in max.

4

This topic was automatically closed 22 days after the last reply. New replies are no longer allowed.