Google Nano Banana Pro 2 and Flux 2 Pro — The gold standard for using reference images, characters or people. Allows for fast and accurate manipulation of images, better than photoshop at generative fill. Excellent for using faces and creating consistent characters and scenes. The use of custom styles and multiple reference images with the ability to mimic any visual aesthetic makes them the best and easiest to use image generators. This also makes all other AI image models such as dalle-3, mystic, and GPT obsolete. Unlimited use on select Freepik plans.

Midjourney — cursejourney's primary tools for creating images. It is the AI image generator that has created the majority of my cursed old photo series and is still one of the best tools for creating images. It can mimic and combine any type of imagery and aesthetic at any aspect ratio. Downsides: Requires learning basic parameters to get the most out of it. It's recommended for those who already have basic image generation experience. May require discord and a separate subscription. Not great at rendering text, not great for editing and manipulating images. 

Stable Diffusion — A truly uncensored option for generating AI images. Excellent for those who want a free option to generate AI images locally on your own computer rather than using a website or other service. Especially good at generating anime style imagery. Can be run via ComfyUI or A1111 or other graphical user interfaces. Downsides: requires a powerful computer with a strong GPU to generate at a decent speed. Requires very high computer knowledge for setting up interface and extensive configuring of models, lora, and parameters to achieve high quality results. The most difficult option that is primarily used for adult content or specialized workflows.