Skip to content
LocalAIBox

Hardware calculator

What hardware do you need to run LocalAIBox?

LocalAIBox runs on hardware you own. Tell us your team size and how capable you want the AI to be, and we'll recommend a GPU, memory, storage, and a rough cost. You buy the hardware; Mako Logics provides and supports the software.

Which platform do you prefer?

Both run the same LocalAIBox software and keep your data on-site. See the difference.

How many people will use it?
How capable should the AI be?

Recommended appliance

NVIDIA RTX 5090

Runs models up to ~32B parameters.

  • 32 GB VRAM
  • 64 GB system RAM
  • 2 TB NVMe SSD
  • Modern multi-core CPU

Estimated hardware cost

$4,000–$6,500

A great fit for most small teams and everyday AI work.

Get an exact quote

These recommendations and cost ranges are non-binding estimates provided for planning purposes only. They are not a quote, offer, or guarantee of price, availability, or performance, and they do not constitute professional advice. Hardware is purchased separately from third-party vendors and actual prices vary. Performance depends on your specific configuration, models, and workload. Mako Logics confirms the exact specification for your needs during your demo. See our Terms of Service.

Two ways to run your box

Windows + NVIDIA or a Mac mini — your choice.

LocalAIBox runs on two appliance platforms. Both keep every prompt and document inside your building — they just trade off in different ways.

Windows + NVIDIA

Maximum speed and the largest models.

  • Best for the fastest responses and many simultaneous users
  • Scales to 70B+ models with professional GPUs
  • Add or upgrade GPUs later as your needs grow
  • A tower — needs more space, power, and cooling
  • Runs the Windows software most offices already manage

Mac mini (Apple Silicon)

Compact, silent, and energy-efficient.

  • Best for small-to-mid teams wanting a tidy, quiet box
  • Runs models up to ~32B on a mini; 70B+ on a Mac Studio
  • The best big-model memory per dollar in a box this size
  • Palm-sized and nearly silent — sips power, sits on a desk
  • Premium, low-maintenance, rock-solid appliance

Choose Mac for a compact, efficient box that fits most teams — or Windows + NVIDIA when you need the most speed, the largest models, or many heavy users at once.

Not sure? We'll recommend one at your demo