Comments

[hidden]

vshock said:

Hello. Sometimes, after waiting a while and trying multiple times, it works. I'm not a paid user.

Then I've no idea how you use it at all. When I hit enter in Imagine, it just opens the price chart, regardless of the settings. Maybe it's restricted by country or platform same as companions, or even age-gated by connecting a verified X account...

[hidden]

Amazing, but how did you get Grok to do it? Last time I used it before they paywalled video gen completely, it wouldn't even agree to do microbikini, they only filtered it more over time. Did they let loose for paid users?

[hidden]

Using the tag "pentagram_cutout" because that's how I've seen the design tagged on other websites, I'll tag some other posts with it (if the AI is able to appropriately make a pentagram instead of a bunch of random lines, that is)

[hidden]

eye-sack-strat-ton said:

That smile, those gorgeous eyes that you just sink into like pools of love.

Oh shit, there's a girl, too. Huh.

I think that tuna is taken, but you know what they say "plenty of fish in the sea".

[hidden]

RikAI said:

The model you use for this image, is it a good one to use for multiple loli images? Sorry for the odd question, been trying to find a loli Ai image generator to use myself, but not sure how to go about finding one.

No problem, though I can't say I'm an expert on this myself.

I think all modern models (Illustrious, NoobAI, NovelAI, anima, and their derivatives) can generate loli without any issues, especially with LoRAs. Use “loli has:metadata” in the search, and you’ll quickly find all the images that include metadata for the generation and the model used - that should give you a useful point of comparison. One thing to note is that NovelAI is a paid online service, while anima generally benefits from having a powerful GPU. Illustrious and NoobAI (both SDXL-based models), on the other hand, will run on pretty much anything.

If you're just getting started with image generation, wai-illustrious-sdxl is a pretty good starting point for almost anything.

For the image above, I specifically used this LoRA; it was responsible for the entire style.
https://civitai.red/models/1884029/blue-archive-cutscene-style-illustrious

Using the metadata from the post together with this LoRA, you should be able to reproduce the image quite easily and generate plenty of similar ones.

[hidden]

The model you use for this image, is it a good one to use for multiple loli images? Sorry for the odd question, been trying to find a loli Ai image generator to use myself, but not sure how to go about finding one.

[hidden]

shadoxart said:

HI ! I love this style so much. Im little ai artist too. Whats IA do you use for make this art ?

Judging by the model tag they used NovelAI, also if you look under the image you will also see a tab called "AI Metadata" where you can see how it was generated.

Always appreciate people sharing metadata.

[hidden]

I don’t think working for others and enduring pain is a good thing. I once said:

While it's nice to be appreciated by others, I won't hurt myself to earn their approval. If everyone could make themselves happy, there would be no need to make others happy. That's why I'm using generative AI.

[hidden]

noppersnop said:

this is why this website is blocked in my country...keep up the good work, citizen

We may sometimes have terrible desires, but as long as we keep them under control, we are citizens. That's why I'm using generative AI.

[hidden]

konny said:

Yes. Initially, I wanted to draw an accurate camera HUD, but in the end I decided not to bother. It wasn't that important to me, and it still looks recognizable as a camera HUD to the average user. I didn't think anyone would actually comment on it, lol. Especially when we're talking about Pixiv.
It was more important to me that Megumin's bandages were always on the correct side, as well as other details.

Tbf, Megumin doesn't know whether Kazuma is recording a video or taking a picture, so I don't see any inconsistency there.

[hidden]

iodoff said:

🤓☝️ umm, actually, that's the kind of viewfinder overlay you'd see on a screen. plus, there's a "rec" indicator, so he's recording a video, not taking photos

Yes. Initially, I wanted to draw an accurate camera HUD, but in the end I decided not to bother. It wasn't that important to me, and it still looks recognizable as a camera HUD to the average user. I didn't think anyone would actually comment on it, lol. Especially when we're talking about Pixiv.
It was more important to me that Megumin's bandages were always on the correct side, as well as other details.

Updated by konny

1 2 3 4 5