The days of having to find reference images of this and that and then compositing that into whatever braindead thought you had is pretty much over.
Similar to needing to settle for the closest approximation in spirit rather than visual accuracy.
At least with AI we can have it all and pretend all the results were intended cause we be "pRoMPt EngInEErs".
The prompt I will be using for all the recommended sites:
A scenic mountain scene, a giant tree with purple leaves growing down in the valley, realistic, the tree is reaching for the clouds
StableCog gives you 100 credits in total but 10 credits per day. Which is quite nice since you can burn them one day and come back and have them grow beyond the "daily quota". Each image also only costs 1 credit so you get the "full amount" in images.
StableCog also has the new Flux.1 model with a decent range of perspectives available. They don't have too many extra's too irritate you with which I find quite nice and straightforward.
Imagine.Art also has the Flux.1 model and some good ones of their own. They however only give you 50 Credits per day. Image generations costs 5 credits so you only have 10 image generations per day available. This resets every 24 hours.
Sometimes the "styles" options come in handy to really tell the stupid botty you want something looking like a sketch or whatnot.
Playground does not offer the Flux.1 model but instead only chooses from their own v1+ type models. It has some ok results but can take more tweaking than others if you have anything slightly specific you want.
They give 10 image generations per day which is not bad either.
Leonardo is one of the first ones I used and have grown pretty large. For the most part I have never cared for how these generators do replacement and other pretty irritatingly bad attempts at editing images.
I don't think there are really generators capable of truly editing an image the way I want it so I just avoid those types of features.
The only thing these generators are good for is to make an initial image and to make it bigger. Simple.
Leonardo gives you 150 Credits, but it forces you to generate 4 images at a time which cost 24 credits in total. So essentially you get about 6 images you can generate per day.
Ideogram is one I forgot about until recently. They were one of the first outside of MidJourney to have proper text reproduction.
They also name things according to their own models but just like Leonardo and them these would be fine-tuned versions of the latest best open models.
Idiogram generates 4 images at a time and these cost 2 credits in total. So you can have 5 generations per day for 20 images in total.
Their site can be a bit janky though and I often am not sure if things are generating but when they do they are quite good. So some patience is advised.
I am adding more AI generators to a bookmark folder since I don't usually need massive amounts or consistency there is not a big reason to stick to one premium account for me.
All in all there are quite a few good free solutions, that for general usage there is no reason to care which one you use.
