Per-Tab AI Model Selection

Feature Requests
1

Summary:

Enable users to assign a specific AI model to each tab within Cursor. Once set, the selected model remains consistent for that tab, regardless of model changes in other tabs.

Advantages:

  • Enhanced Workflow Efficiency: Allows developers to tailor AI assistance based on the specific needs of each task or file, facilitating a more streamlined and efficient coding process.

  • Reduced Context Switching: Minimizes the need to manually switch models when navigating between different tasks, thereby reducing cognitive load.

  • Improved Task Management: Supports parallel workflows by enabling different AI models to operate simultaneously in separate tabs, each optimized for its respective task.

Current Workaround:

At present, users must manually switch the AI model each time they change tabs, as the model selection is global across all tabs.

Other Cases:

6 Likes
2

Massive :plus: to this

I frequently have 2-3 tabs with the same initial prompt using different models to test which might be best at a task, and I would love the option to keep each conversation going with the initially-selected model without needing to remember to change the selection between tabs.

3

Really surprised this feature is not more in demand. I frequently am working with two different branches concurrently and need different models for different jobs. The fact that changing the model in one project affects the other project makes zero sense and makes it impossible to do these types of projects concurrently.

1 Like
4

Yes, it’s really annoying that on different tabs, it keeps the last selected model from a previous prompt in another tab. It makes no sense. This should be a default feature

5

+1!

Actually not only tabs, I have multiple cursor windows opened while working on multiple projects, some of them are open source where I use cheap models and on work projects I often use max, and it’s pain to constantly switching back and forth

1 Like
6

Hi Cursor team,

Thanks for building such an excellent tool — it’s been a huge boost to my workflow.

I’d like to request a feature that would make working with multiple tabs even more flexible: per-tab model selection.

Current Behavior

Right now, when I change the model (e.g., o3, Claude 4, etc.) in one tab, it changes the model for all other tabs as well.

Why This Matters

Often, I’m working on multiple tasks at once:

  • One tab might just need a Claude 4 implementation.
  • Another might need deeper reasoning for planning with o3.

Switching models for one tab unintentionally affects my other work, forcing unnecessary toggling and reducing context consistency.

Proposed Feature

  • Make the selected model specific to each tab.
  • Optionally include a “Global model lock” toggle for those who prefer synchronized behavior.

This would allow a more contextual and modular workflow — especially helpful when juggling different projects or types of tasks at once.

Thanks again for your hard work and consideration!

7

Currently, if I have multiple chat tabs and change the model in one, it changes the model in all chat tabs. I would like to be able to chat using a given model in one tab, while keeping the chosen model constant in other tabs.

8

Being able to have parallel tabs or sessions working with different models would be great. Currently, the last manual model change applies to all chat tabs and cursor windows.

9

I believe this was solved in Changelog - Aug 6, 2025 | Cursor - The AI Code Editor | Cursor - The AI Code Editor (Separate models per Agent)

2 Likes