I'm Tomasz. I make AI images with FLUX in ComfyUI on an RTX 5060 Ti 16 GB, most days, as a side project next to a full-time job.
The question I kept asking — and kept seeing on Reddit — was always the same: will this model run on my GPU, and which file do I download? The answers were spread across a hundred threads, half of them out of date. So I built the page I wanted to find.
What makes this different
- Real file sizes. Read from Hugging Face to the byte, and re-checked.
- One visible rule. The same calculation for every GPU and model, explained on one page.
- Labels on every number. Calculated, measured or reported — you always know which.
- Laptops included, with their real VRAM, not the desktop card's.
How it was built
The data collection — reading file listings and spec sheets — was done with AI help. I decided what to collect, how to judge it and how it is presented. Every figure links to its source so you can check it, and I fix mistakes as soon as someone points them out.
Money
This site is free and has no ads. When a model does not fit well on your card, some pages show one clearly marked affiliate link to Vast.ai, where you can rent a bigger GPU by the hour. If you rent through it, I get a small share of what you spend; it costs you nothing extra. It never changes a verdict: the same rule decides every page.