@david_nix Every AI model I use in comfyui nearly maxes out my 64gb of ram. In my opinion start with 64 and if the price drops in the future buy the same 64gb kit to make 128gb.
@arena@xai@elonmusk I’m sure it’s good. But I’ll never pay for Grok again after all the issues of hitting limits and being unavailable. Open source LTX 2.3 I2V locally is doing everything I need.
@TheAhmadOsman I cancelled my Kling subscription and have been using LTX 2.3 locally in Comfyui with RTX 4090. Cost wise is about the same with power bill but I’m getting nearly 4 - 5x the amount of generations at the same cost as I did with Kling. But the main focus for me is it’s all local
@WillieDJenkins1@tzukima No such thing as too much haha. I couldn’t be happier with the board. I’ve used 4 different kinds of ram in it and they all have worked with their profiles. And gigabyte is still pumping out bios updates.