Promptus generates locally, so what you can make depends on your GPU. It runs across a wide range of hardware — from a GTX 1060 to an RTX 5090, and on Apple Silicon. Here is exactly what runs at each tier.
Every tier generates locally. The tier decides how big a model you can load and how fast it finishes.
NVIDIA GTX 1060 6 GB or better
Or CPU-only, which works but is slow
NVIDIA RTX 3060 12 GB or better
Or Apple M2 / M3 with 16 GB unified memory
NVIDIA RTX 4070 or above, up to RTX 5090
Or Apple M3 Max with 36 GB or more
Apple M1 or later, 8 GB unified memory
macOS 12 Monterey or later
No compatible GPU? Promptus also supports cloud-backed generation through API routing. It is slower than local, but it works — and it is the right option while you wait on hardware.
AMD cards: Radeon support is on the roadmap, not shipped. If you have an AMD GPU today, use CPU-only mode or the cloud plan.
Tap a question to see the answer.
One payment of $49 and Promptus is yours — unlimited local generation, no subscription, no per-image fees.
Not sure your setup qualifies? Email support@promptus.ai with your specs and we will tell you exactly what will run.