similar to being able to use local LLMs for chat it would be great to be able to use local image generation AI models as well, e.g. stable diffusion.
in fact some of the suggested image generation models in the DT4 settings are open source models that can run locally yet using replicant.com is a requirement…
There are already several apps to perform local image generation (and having an endless amount of options), e.g. DiffusionBee or Draw Things. What would be the real advantage here?
doesn’t your question hold true for any image generation within DT4, ie for local as well as cloud based?
I am using Draw Things with local models today as I used to use Anything LLM prior to the beta of DT4 coming out, which is now my primary interface to my local (and cloud) LLMs. Similarly it would be convenient to have DT4 replace Draw Things as my front end to image generation as the output most of the time ends up in DEVONthink anyways. I even tinker around the prompts in DEVONthink…so I guess my answer to your question is convenience as well as one stop shop…