Black Forest Labs Knocks it Out of the Park with Flux.1 Dev

This is probably the best image production model I've worked with on my rig. I threw at it my usual "Strawberry frog in a cranberry bog in the fog on a log" prompt. I also gave it the wizard in his lair prompt, but this time I gave it a young wizard looking back at the camera with an embarrassed smile, and it absolutely nailed it on the first try. Salarymen surfing the Shibuya crossing... Devil and angel battle... Adventure mouse... One prompt involved street in Palermo covered with spaghetti that is surprisingly good. I absolutely love this model. It doesn't always nail text, but if the text is short enough and in a reasonable context, you can usually get something believable in a few tries.

It does not do well with prompts involving vector, painting or other forms of art yet. It's amazing with photoreal and stuff that looks like 3D render.

Well done Black Forest Labs!