Nano Banana 2 on tensor.art's new canvas feature (which doesn't seem to allow links and only half-follows the prompt). Prompt below the fold:
Not today Justin

oozey mess
One Nice Bug Per Day

Product Placement

shark vs the universe
Claire Keane
hello vonnie
almost home

pixel skylines
todays bird
Sade Olutola

PR's Tumblrdome
d e v o n

Love Begins
$LAYYYTER
Aqua Utopia|海の底で記憶を紡ぐ

Kiana Khansmith
i don't do bad sauce passes
No title available
Xuebing Du
seen from United States

seen from United States
seen from Taiwan
seen from United Kingdom
seen from France
seen from United States
seen from Barbados

seen from Maldives
seen from Tunisia

seen from France
seen from United States

seen from Indonesia
seen from United States
seen from United Kingdom
seen from United Kingdom
seen from United States

seen from Singapore

seen from Malaysia

seen from T1
seen from Brazil
@fn205ai
Nano Banana 2 on tensor.art's new canvas feature (which doesn't seem to allow links and only half-follows the prompt). Prompt below the fold:
Link. Z-Image-Turbo on tensor.art, with Z-Image Film-Texture-movie LORA. Prompt below the fold:
Link: Z-Image-Turbo on tensor.art. Prompt below the fold.
No link bc nsfw. Z-Image on Seaart.ai with the Amateur Photography Lora. Prompt below the fold
Link. Z-Image on tensor.art. Prompt below.
Building better (fetish) AI prompts
One of my big problems in AI image generation (for fetish purposes) is that it can be hard to get the prompt right, because it's not what the generator is designed for. I end up building Jenga tower prompts, where I try a bunch of ideas, tweak until I get it just right, and then only nibble at the edges when revising because I don't want to break it and have to start all over again.
If you've been looking at my prompts, and even my pictures, I'm sure you see the evidence of these points where I don't have any new ideas and just keep recycling the same core. One reason I share all my prompts is so that you, the reader, can think of how you'd like to tweak what I made: both to suit your particular tastes and to go beyond my limited idea space.
Sometimes prompt enhancers help, especially for coming up with revised art styles that can give an exciting new look to build on. I've been having pretty good luck with a free one for Z-Image prompts (my current favorite image model): https://aiomnigen.com/tools/prompt-generator
Below the fold, I have some comparisons of the prompts it designed, with images generated from the different prompts.
parameter link. Z-Image on krea.ai. Prompt below the fold:
parameter link. Z-Image on krea.ai. Prompt below the fold:
parameter link. Z-Image on krea.ai. Prompt below the fold.
parameter link. Z-Image on krea.ai. Prompt below the fold:
parameter link. Z Image on krea.ai. Prompt below the fold:
Link. Z-Image Turbo with Full Round_ LORA on seaart.ai. Prompt below the fold:
Link. Z-Image Turbo on seaart.ai. Prompt below the fold.
Link. Z-Image Base, with Amateur Photography LORA on seaart.ai. Prompt below the fold.
No link bc NSFW. Z-Image Base on seaart.ai. Prompt below the fold:
[LINK] Z-Image Base on seaart.ai. Prompt below the fold:
Link. Z-Image on tensor.art. Prompt below.