RAM is the spec people most reliably get wrong — buying 64 GB to browse the web, or clinging to 8 GB while wondering why their laptop chokes on a dozen tabs. The right amount isn't about your budget; it's about what you actually do all day. Here's the honest map.
The task-by-task table
| What you do | RAM | Why |
|---|---|---|
| Web, email, streaming, office | 16 GB | 8 GB technically works but modern browsers and apps make 16 GB the comfortable floor. |
| Heavy multitasking, many tabs | 16–32 GB | Dozens of tabs plus chat apps and a video call add up fast. |
| Photo editing, light coding | 32 GB | Large images, IDEs, containers, and previews all live in RAM. |
| 4K video editing, VMs, big data | 32–64 GB | Timelines, scratch caches, and virtual machines are genuinely hungry. |
| Local AI with RAM offload | 32–64 GB+ | Offloading model layers that don't fit in VRAM uses system RAM directly. |
Capacity beats speed
RAM has two numbers: how much (GB) and how fast (MHz / DDR generation). For general use, capacity wins every time — running out of RAM forces your system to swap to disk, which is dramatically slower than any speed difference between RAM kits. Get enough gigabytes first; chase faster RAM only if you're a gamer pairing it with a CPU that benefits (some do, noticeably).
The local-AI footnote
If you dabble in local AI, RAM does double duty: when a model is too big for your GPU's VRAM, the overflow spills into system RAM (offloading). It's slower than VRAM, but 32–64 GB of RAM is what lets a modest GPU run a model it otherwise couldn't load at all. If that's on your roadmap, lean toward 32 GB minimum.
FAQ
We may partner with companies or groups to affiliate hardware products based on user needs, earning a commission from qualifying purchases. Recommendations are general guidance and vary by specific applications and workflow. Data current as of June 2026.