I Built a Fast Custom Gaming Laptop Nearly From Scratch With the Framework 16

When ordering your Framework 16, you can choose between the AMD Ryzen AI 7 350 processor to start ($1,799 pre-built with Windows 11) or the faster AI 9 HX 370 ($2,549 pre-built), with a midrange Ryzen AI 7 350 option in between. The pre-built configurations start with 16GB of memory, a 512GB solid-state drive, and a power adapter. (However, Framework includes just six USB-C expansion cards in these pre-built configurations.)

(Credit: Joseph Maldonado)

The removable Nvidia GeForce RTX 5070 GPU attachment option, which replaces the expansion bay shell, is available for any pre-built configuration for an additional $650.

The DIY model is less expensive than the pre-built version, starting at $1,499 for the Ryzen AI 7 350 platform. This is as close to a bare kit as possible, with no expansion cards, memory, storage, operating system, or even a power adapter included—it assumes you will bring all these components on your own. You can add one or all of these parts to your order, including a Windows 11 install key if you don’t want to run a Linux distro. 

(Credit: Matthew Buzzi)

The same applies to memory and Framework-specific extras, such as expansion cards, spacers, extra input modules, and more. Upgrading from the base version to the Ryzen 9 AI HX 370 processor costs an additional $300, and the graphics upgrade costs $650 as a DIY upgrade, too.

For pricing my exact review unit, let’s stick to what I installed and consider it part of the product. For expansion cards, we received two USB-C cards, two USB-A cards, an Ethernet card, a DisplayPort card, a 3.5mm audio card, and an SD expansion card. (These range from $11 to $39 each.) I also installed the 32GB of memory (two DDR5 16GB sticks) and a 1TB SSD that I received, and Framework finally sent a few of the keyboard deck spacers, a lavender bezel, and the optional number pad extension.

(Credit: Matthew Buzzi)

Adding those to the RTX 5070 graphics module, a Windows 11 key, and the power adapter, our review configuration would run you a steep $3,198. Framework sent a few other extras, including another 2TB SSD, but since all memory and storage expansion is totally optional for the user, I left those out of the final test configuration. The final test unit represents a realistic build that a power user may order and assemble themselves, with some optional spacers and add-ons to swap in or out as needed.

(Credit: Joseph Maldonado)

Regardless of which model you choose, the upgradable GPU module is particularly exciting, fulfilling Framework’s promise of a completely upgradable laptop. Framework’s solution is surprisingly elegant but not exactly cheap. (Then again, has buying a new graphics processor ever been?) At least you can upgrade just the GPU if you want to in the future, without throwing out your whole laptop. Let’s see how well it performs.