Anyone here actually used Getimg.ai regularly? Saw a pretty detailed review and wanted to ask about it.

My company occasionally needs quick concept art, so I’ve been looking for a hassle-free online image generator. I came across a detailed 2023 review of Getimg.ai that was pretty thorough, and I want to check with people who’ve actually used it.

The review said it’s built on Stable Diffusion, so you can use all kinds of SD models for different styles, plus it has advanced features like inpainting and ControlNet. The author praised it for saving all the hassle of setting up WebUI locally, and said it’s beginner-friendly for people who’ve never touched those tools.

Feature-wise, it’s pretty complete: text-to-image, image-to-image, DreamBooth to train your own model, inpainting, outpainting, and even Instruct Pix2Pix for editing images with natural language. The author mentioned that other sites charge over thirty bucks to train one DreamBooth model, but this one’s cheapest tier lets you train two.

The complaints were legit too: the outpainting tab looks nice but is clunky to use—drag it into a non-square shape and you get distortion. The face enhancement feature is basically abandoned. But the image editor’s blending and background swapping are actually decent.

Pricing is per image with credits, from $12 for 3,000 images up to $99 for 60,000. The author said 60,000 is already industrial-level usage.

So, anyone still using it? Has it kept up with updates over the past two years?

The thing about SD is that it’s built on a solid foundation, so you’ve got way more model options—that was its main selling point back in the day.

Yeah, I’ve run into that issue too—when you stretch it to a non-square aspect ratio, it gets all distorted.

Yeah, nobody uses face enhancement anymore, it’s been replaced by way better solutions.

DreamBooth is legit cheaper for training two models compared to other services.

Paying per credit to generate images feels great, but my wallet can’t take it.

Yeah, it’s definitely beginner-friendly, saves you from having to install WebUI.

This is a review from 2023, the features should have been iterated a lot by now.

The background swap feature was pretty decent from what I remember.

+1 on the distortion issue when you stretch it into a non-square shape. I usually just play it safe and generate in square format, then crop it afterwards—saves me the headache.