How to Remove or Disable GPT-5 in Cursor's Model List

8/11/2025

Can You Really Remove GPT-5 From Cursor's Model List? Here's the Real Deal

Hey everyone, so you're digging into Cursor, getting everything set up just right, & you start looking at the list of available AI models. It’s packed with options from OpenAI, Anthropic, Google, & more. But maybe you have your favorites, or maybe you just want to clean up the interface a bit. A common question that pops up is: "How do I remove a model I don't use, like GPT-5, from the list?"

It's a totally valid question. A cleaner workspace is a more focused workspace. But here's the thing, & I'm just going to be straight with you: you can't actually permanently delete a default model like GPT-5 from Cursor's settings panel.

I know, that might not be the answer you were looking for. Turns out, the list of pre-configured models is baked into the application itself. Think of them as part of the core package. The developers provide this extensive list to give users a wide array of choices right out of the box. So, while you can add your own custom models, you can't go in & start deleting the built-in ones.

But don't worry, that doesn't mean you're stuck with a cluttered list in your chat window. The way Cursor handles this is through enabling & disabling models. It's a subtle but important difference. You're not removing it, you're just hiding it from your active workspace. Honestly, it accomplishes the same goal for your day-to-day workflow.

Let's walk through how to actually do it, plus a few other things you should know about managing models in Cursor.

How to "Remove" a Model by Disabling It (The Right Way)

So, you want GPT-5 gone from your dropdown list. No problem. The process is super simple & just takes a few clicks.

1. Open Your Cursor Settings: First things first, you need to get to the settings panel. If you're on the main screen, look for the little cogwheel icon. It's usually up near the top. Click on that. This is your command center for everything related to configuring Cursor.

2. Navigate to the "Models" Section: Once you're in the settings, you'll see a menu on the side with different categories like "Account," "Keyboard Shortcuts," & so on. You're looking for the Models tab. Click on it.

3. Find & Uncheck the Model: Here it is—the master list of every model Cursor has pre-installed. You'll see a long, scrollable list with names like Claude 3.5 Sonnet, Gemini 2.5 Pro, & of course, the GPT family.

Scroll down until you find OpenAI GPT-5. You'll see it has a little checkmark next to it, which means it's "enabled" & will show up in your chat model selector.

To "remove" it, just click the checkmark to uncheck it.

That’s literally it.

Now, when you go back to your chat window & click the button to switch models, GPT-5 will be gone from the list. It's still in your main settings panel if you ever change your mind & want to bring it back, but it's no longer cluttering your active choices. Pretty cool, right?

You can do this for any other model in the list. I personally like to disable the ones I know I'll never use, just to keep things tidy. It makes switching between my go-to models, like Opus or Sonnet, that much faster.

Why You Can't Actually Delete It

So why does Cursor do it this way? Honestly, it makes sense from a software design perspective. The models list is part of the core feature set. By allowing you to enable or disable them, they give you control over your workspace without letting you accidentally break something by deleting a core component.

It also means that when Cursor pushes an update, it can manage the list of available models, maybe adding new ones or updating existing ones, without messing up a user's custom configuration. It's a simple, non-destructive way to handle user preferences.

What About Adding Your Own Models?

Now, while you can't remove default models, you absolutely CAN add your own. This is where Cursor gets REALLY powerful, especially for developers who might be working with specialized or local models.

Let's say you're running a model locally on your machine using something like Ollama. You can actually pipe that directly into Cursor. Here’s the gist of how it works:

Run Your Local Model: You'd get Ollama set up & pull the model you want to run (e.g., Llama 3, Mistral, etc.).
Expose it on a Local Server: Ollama does this for you automatically, usually on a specific port on your machine.
Add it in Cursor: Back in that same Settings > Models panel, you'd scroll all the way to the bottom. There's an "Add Model" button. You'd give your model a name, and then you'd need to configure the endpoint. This usually involves overriding the OpenAI Base URL to point to your local server address.

This is a game-changer because you can use completely open-source models, or even fine-tuned custom models, directly within your IDE, for free. You get all the awesome features of Cursor's interface with the power & privacy of a model running on your own hardware.

The Power of AI Chatbots in Modern Business

Thinking about all this model integration brings up a bigger point about how businesses are using AI these days. It's not just developers in their IDEs. AI, especially conversational AI, is showing up everywhere.

Take customer service, for example. For years, the options were either a human support team (which is expensive & not always available 24/7) or a really basic, frustrating chatbot that could only answer a few pre-programmed questions.

Now, things are different. This is where platforms like Arsturn come into play. Businesses are using it to build their own custom AI chatbots, trained on their own data. Imagine having a support agent on your website that knows everything about your products, your documentation, your company policies, & can provide instant, accurate answers to customer questions anytime, day or night. That’s what’s possible now.

Instead of a clunky bot that says "I don't understand," you get a seamless experience that feels like you're talking to an expert. This is HUGE for customer engagement & satisfaction. It frees up human agents to handle the really complex, high-touch issues, while the AI handles the frontline questions.

It's pretty much the same principle as adding a local model to Cursor. You're taking the power of these advanced language models & training them on a specific set of knowledge to perform a specialized task. For developers, that task is coding. For a business using Arsturn, that task is providing incredible customer support, generating leads, & engaging with website visitors in a meaningful way. It's all about creating personalized, intelligent experiences.

Tying It All Together

So, to wrap things up, you can't technically delete GPT-5 from Cursor's model list, but you can easily disable it to hide it from your active workspace, which is just as good for day-to-day use. Just head to Settings > Models & uncheck the box.

And while you're thinking about how you customize your own tools, it's worth considering how that same idea applies to business. The ability to create custom, data-trained AI assistants is fundamentally changing how companies interact with their customers. It's a super exciting space to watch.

Hope this was helpful! Let me know if you have any other questions or if you've found any cool custom models to run with Cursor.