Currently I'm using both of your standalone offline tools. I have been generating material in Aiimages over the last half a year, and this month have started using InvokeAi mainly for the Unified Canvas feature. Taking images made in Aiimages, and importing them to InvokeAI to develop them further with the canvas toolset.
I have three instances/installs of Aiimages that I use for separate content/workflows using different models, and also because after 20k+ images are in history, in my experience it gets a little slow and frequently crashy when scrolling through the vast galleries.
With InvokeAi, I find its design more of a hassle for accessing earlier generations because it only loads 50 (or is it 100?) images in to cache, and then I have to constantly scroll down and click to load another 50/100 images. I generate LOTS of images per prompt so accessing the images becomes cumbersome in InvokeAi. Very klunky and limiting in contrast with Aiimages, in that respect. Pros and Cons yo!
For my workflow, being able to load up one of the Aiimages instances and quickly scrolling the history to get the precise prompt for an image to work on inside InvokeAi's unified canvas has been invaluable. So while I still have the images, and still have the dream_md / .txt prompt histories for each of the three Aiimages instances intact in their respective folders, I can no longer easily scroll the history to access the image with corresponding prompts.
Aiimages handles the Style/Content fields a little differently to InvokeAi with its Prompt/Negative fields. For instance, if I copy a Style/Content prompt from Aiimages into InvokeAI's prompt field, all is well until I reload the prompt at a later stage, as it then takes the original prompting character/token sequence and populates the Negative prompt field with any [negatives] from the original Aiimages prompt. It appears to modify the generated output aesthetic quite dramatically at times when this happens, so I usually just copy the prompt from Aiimages back in to the InvokeAi prompt field each time I want to return to a theme (say for going past the 1024 x 1024 limit in Aiimages) to keep the precise prompt.
If I'm not mistaken, this also means the metadata when taking an image from one application to the other changes the numerical(/character?) string that the prompts are converted into, so I very much hope to retain the history of images generated in my Aiimages instances over the last 6 months inside Aiimages for easy access of/reference to the original prompts.
Thanks so much for releasing these tools for the beginners, the Unity devs in the case of Aiimages, and for the folk who just like to run their machines offline to avoid the overlords snooping and auto-update tampering with their delicate work environments! :D x