Flux seems to give the backgrounds even more detail and coherence compared to MJ, it's a surprisingly (or maybe not given its size) great model.

where can we try/use? Is it just the obvious google result of https://getimg.ai/models/flux ?

Replicate, although not fully free, has the bonus of outputs from Flux (from their endpoints only) able to be used for commercial purposes whereas that is normally only applicable for Flux schnell:

https://replicate.com/black-forest-labs/flux-dev

https://replicate.com/black-forest-labs/flux-schnell

Are you sure that's true?

https://github.com/black-forest-labs/flux/blob/main/model_li...

From the license: "We claim no ownership rights in and to the Outputs. You are solely responsible for the Outputs you generate and their subsequent uses in accordance with this License. You may use Output for any purpose (including for commercial purposes), except as expressly prohibited herein. You may not use the Output to train, fine-tune or distill a model that is competitive with the FLUX.1 [dev] Model."

This license seems to indicate that the images from the dev model CAN be used for commercial purposes outside of using those images to train derivative models. It would be a little weird to me that they'd allow you to use FluxDev images for commercial purposes IF AND ONLY IF the model host was Replicate.

Replicate has a special agreement outside the license, apparently, see the licensing: https://replicate.com/blog/fine-tune-flux

They also now just added a "Commercial friendly" tag to the endpoints above.

Yes, it's weird.

https://www.youtube.com/watch?v=QYVucud3ptc has good coverage of running locally and via web (https://fal.ai/models)

Their Github repo provides some links if you want to try it online: https://github.com/black-forest-labs/flux

Grok on Twitter Premium has Flux integrated.

MJ has always been this pixar-style-hyperfit SD variant. It almost seemed to just repeat google images but with gloss.

Or at least a year ago it was like that.

That is the base style. You can get much different image qualities adding different types of film, cameras, etc to the prompt.

With that said, I loved Midjourney a year ago but I am at the point I have seen enough AI art to last several lives.

AI art reminds me of eating wasabi. It is absolutely amazing at first but I quickly get totally sick of it.

Here's a project I did over a year ago in Midjourney where I took all art for the card game Spectromancer and re-themed each piece in 26 different styles

https://kellydornhaus.github.io/Variations/

For a period I daily checked its best outputs page: the "pixar" style was frequent, but far from being the only one. "Typical" like 10 is the type (mode) in 10+9+9+8+8+7+7+6+6+5+5+4+4+3+3+2+2+1+1 - but still 10% of the whole and just one possibility of all.

Midlibrary.io recognizes 5500 styles Midjourney knows.