Yo, FLUX.2 just dropped—it’s an open-source AI image gen and editing model. A ton of people were already messing with the original FLUX, and now they’re straight-up on version 2. I’m keeping an eye on the open-source side 'cause I wanna run it locally for some automated testing images.
Calling commercial model APIs gets pricey real quick when you’re doing a bunch of requests. Open-source means you can deploy it yourself, stress-test it, and mess around however you want.
But I’m wondering how hungry FLUX.2 is for VRAM—can a consumer-grade card handle it? Anyone got it up and running yet? How’s the deployment going, and is it a hardware hog?