RAM is expensive in phones, too – so how much do you really need?

Follow ZDNET: Add us as a favorite resource on Google.

Highlights taken by ZDNET

• Phones have more RAM than ever, while prices are increasing.
• Flagship phones have more RAM that most people need.
• Most people only need 8GB or 12GB.

If you’ve run through the specs of new phones while shopping, you’ve probably seen high RAM prices among the most recommended features. Phones are getting more and more random access memory (RAM), but how much does your phone really need?

There are two metrics advertised in GB: RAM and storage. Storage means how much space you have to hold apps, photos, videos, documents. RAM, on the other hand, refers to temporary memory and represents how much the phone can do at once. In short, its performance.

Also: I’ve been using Samsung foldable phones for 7 years, and the Z Fold 8 Ultra is finally good.

But there’s only so much your phone can do, anyway. While more RAM is better in theory, more phone comes with a price — literally. Thanks to AI data centers are collecting almost all available RAM (it is estimated that data centers will demand 70% of memory chips in 2026), prices are more expensive than ever and continue to rise.

Phone manufacturers, in turn, pass this increase on to customers. Google has already announced that the Pixel 11 series will be more expensive due to this issue, and we will likely see higher priced iPhones this fall.

Given the soaring prices, you might be wondering how much RAM your phone really needs. Here’s a look at how much you need depending on what you do with your phone. And yes, you can have too much.

How much is too much?

6GB of RAM or less

It’s getting harder to find phones with only 6GB of RAM, but some budget-friendly, entry-level models start here. If you’re considering a phone with only 6GB, know that it can handle basic tasks like texting, surfing, and web browsing, but you’ll be closing background apps all the time.

Who needs 6GB of RAM: If you want a very basic phone that can call and text but doesn’t need a lot of apps, 6GB should be fine. You may notice apps reloading when you switch frequently.

8GB of RAM

This is the most common memory level today. A phone with 8GB of RAM will run multiple apps smoothly, let you stream without stuttering, switch between apps without a problem, run multiple apps at once, and even play with AI like Gemini.

Also: Samsung Galaxy Z Fold 8 vs. Z Flip 8: How Samsung’s two new flips compare

The tradeoff is that since most phones include AI like Apple Intelligence or Google Gemini, your usable RAM isn’t as much as you think, so you may find your phone shutting down apps to prioritize tasks. As AI becomes more resource hungry, this may start to become a big problem.

Who needs 8 GB of RAM: If you stick to multitasking between apps but don’t edit photos and easily use AI features, 8GB will be enough.

12GB of RAM

Most of the top phones today like the latest Google Pixel, Apple iPhone, or Samsung Galaxy land with 12GB of RAM. While this is enough to handle the powerful on-board processing required by AI features, keeping multiple apps open constantly, editing photos and video, and being efficient most of the time, it’s more RAM than most people really need. Of course, you’ll be paying a premium for this much RAM, so you have to decide if it’s worth the cost.

Who needs 12 GB of RAM: If you like to take advantage of everything your phone can do and use your phone instead of simple work that you usually do on a laptop or desktop, you want 12GB.

16GB of RAM

Here, you’ll find premium models like the Google Pixel 10 Pro XL, the Samsung Galaxy Z Fold 8 Ultra and the S25 Ultra, and the Motorola Razr Ultra. 16GB of RAM makes for a very powerful device, but it’s more than most people need. With this much RAM, you can edit 4K or 8K raw video, play games, run local AI models, and handle many tasks with ease.

Who needs 16GB of RAM:You only need this much RAM if you’re a real power user who keeps a dozen apps running at once, want a desktop-style experience, or want to future-proof extended AI models down the line.