sebba

@thecosmos Please add a PDF upload feature and support for saving posts from X 🙏

@lloydcreates Wow ! Thank you so much <333

@1sebba love this

@Affinity add Recraft v4 vector for vector generation pls

Your workflow, your call. Affinity now gives you more choice over AI models, including Nano Banana 2, across Generate Image & Generative Edit 🍌 If you have Canva premium, you can unlock Canva AI tools directly in Affinity for a supercharged workflow: affinity.studio/canva-integrat…

MJ 8.1

The Hermes Agent Creative Hackathon sponsored by @Kimi_Moonshot has ended! Finalists were selected by Nous and Kimi staff out of 227 submissions on creativity, usefulness and presentation. We were absolutely blown away by the creativity of the things you all built using Hermes. It was tough choosing winners, huge thank you to all who participated! Winners below:

I really want to try moodboards in Krea. I have so many moodboards that just never worked properly in MJ.

@g_rant__ @krea_ai Wow !

some images i’ve generated today with Krea 2 comment below & i’ll dm you an access code :)

@ibexdream @ciguleva I’m just judging by my own preferences, but I don’t like that it’s not something personalized, and it’s not always clear how it blends with other styles. I also don’t like that it locks in the color definition

@ciguleva Yeah, shocking, I know. It was the same for me, especially since I post srefs regularly...

I mean... with sref codes, everything is possible

🚙

@JCKHLRY Они работают над тем чтобы это исправить, будем надеется что это не займет у них много времени

GPT2 keeps over-graining my images when I edit, I surely can't be the only one?!

@The_Sycomore @jameygannon Thank you so much <3

No, I specified the type of content I want on each page to avoid having unnecessary pages for exploration (fonts, manifesto, etc.). The prompt: Create a polished multi-page brand bento kit as multiple images for (ref-logo) The brand identity should be inspired by (ref MJ1, ref MJ2, Ref MJ3, Ref MJ4) Page 1: Main logo Page 2: Logo variations Page 3: Icon set Page 4: Brand baseline only, with illustration-inspired visuals in a more abstract direction Page 5: Product mockups Page 6: Kidswear mockups, shown through lifestyle photos with flash photography Page 7: Examples of urban campaign ads mixing illustration and children Page 8: Example app interface views Page 9: Social media visuals

Inspired by @jameygannon's post, I tried to see if it was possible to create a brand by using different models only for what they’re best at. Recraft V4 for the logo, Midjourney for the aesthetic, GPT-Image-2 for the kit . And… it’s surprising.

cmptr Midjourney v8.1

@ciguleva Это была шутка 😭

@EYESTR4IN @lemonavenuee ::3 ❤️

@1sebba @lemonavenuee Holy this is so good

Niji 7 ::3

logo -> claude semiotician -> system x.com/agua90era/stat…

What annoys me most is this lack of critical thinking. Instead of judging the model in the context of its current stage, some people criticize it as if they were looking at a full release. Personally, I think the model is already very interesting and has huge creative potential.

I really do not understand how people can bash Midjourney’s alpha model as if it were already a finished product. It literally says everywhere that this is an alpha - even on the generation website itself. Is it really that hard to understand that this is not the final version.

working on adding decent bitmap font support to my crt shader in threejs / webgpu