Description
Image imported from Derpibooru.
Original: https://derpibooru.org/images/3413885
Metadata at the time of import:
Original: https://derpibooru.org/images/3413885
Metadata at the time of import:
| Field | Data |
|---|---|
| Favorites | 67 |
| Upvotes | 102 |
| Downvotes | 7 |
| Score | 95 |
| Comments | 4 |
| Uploader | Button Fixer |
Original description below this line
First image I made with Flux AI. If it wasn’t for the computational resources it requires I imagine it would suceed Stable Diffusion in the long term.
I hope for a new model as well, but I don’t think it’ll happen as all the bigg companies are so worried about bad pr that a really finetunable model is unlikely. I want to be proven wrong though.
Yes, this is a completely different level. I have been waiting for a long time for a model to appear in which the prompt is written without using key tags that are loosely related to each other. Last week I came across auroflow that uses a similar prompt principle, but in my opinion the model is a failure. I really hope that the corrected SD3 will be even better than flux and we will see v7 on something fundamentally new and different from SDXL sd2 and sd.15
I like your Luna video. I keep trying with local video but I’m not getting anywhere so far.