Anything2real (Qwen)
https://redd.it/1p09siu
@rStableDiffusion

LocalGen is released. Now you can run SDXL locally on your iPhone.

https://redd.it/1p0egf8
@rStableDiffusion

Looks like Flux-2 will be available at a hackathon in SF this weekend

Hackathon is noscriptd "FLUX: Beyond One"

Quote from the BFL hackathon website: "Black Forest Labs is launching something big, and you're invited to build with it first."

https://cerebralvalley.ai/e/bfl-hackathon

But hold your horses:

1. Public launch could ofcourse be weeks later
2. Whether an open weights variant of Flux-2 will be released is uncertain

PS. BFL advertised this hackathon and that website on their official Twitter account (just saying because there's all sorts of weird fakes out there these days)

https://redd.it/1p0ptvk
@rStableDiffusion

VHS tapes restore using AI

https://redd.it/1p0t4kk
@rStableDiffusion

Illustrious Flowmatch

https://redd.it/1p11gr5
@rStableDiffusion

🚀 [RELEASE] MegaWorkflow V1 — The Ultimate All-In-One ComfyUI Pipeline (Wan Animate 2.2 + SeedVR2 + Qwen Image/Edit + FlashVSR + Painter + T2V/I2V + First/Last Frame)
https://redd.it/1p11ur9
@rStableDiffusion

Flux Kontext is great, is there anything better?

Hi there!

I am trying to create an avatar out of a picture (just one selfie), and it does work quite nicely.



but I wonder if there is an even better model out there.

Any recommendation for a one picture avatar generation?

Also, any recommendation for prompts in different styles?

Thank you and have a great one!

https://redd.it/1p12k5e
@rStableDiffusion

Nostalgic return to SDXL, model merge, alpha mask/IPadapter mix-up, one lora, WF in comments.

https://redd.it/1p15v0x
@rStableDiffusion