just for my information, since I will buy a new PC in the next days: this "16Gb" model, can run in local on a 16Gb Gram GPU card? For example a 3070 with 16Gb?

Does this file help to save space in terms of VRAM consumption?

I'm unable to use those "all in one" checkpoints in ComfyUI. I know I must put this checkpoint in the checkpoint folder, but my working workflow (that I'm used to work with) uses the separate model and clip files that are located in different folders. When I try the workflows embedded in the image gallery below, they have lots of undefined nodes.

So how different is this version compared to the 11GB version available specially in term of optimization, and is this F8 or F16 ?

nice.

Can you make dev 8bit with t5xx 16 bit?

Because t5xx 8bit is degrading badly images

what the difference between this version and this one " flux1-dev-fp8-all-in-one.safetensors "?..

THE BEST!

works great in Ruined Fooocus ! thanks a lot

не работает в Fooocus(((