Join The Tinkerer on Whop to download this model and many more — all for one monthly payment. Plus, get early releases, private tools and a lot more. 👉 Join on Whop
Recommended Settings Forge Neo
Sampler: Euler a
Steps: 20–30
CFG: 6-7
Resolution: 896×1152 or 832×1216 (portrait-leaning works best)
VAE: Baked-in
Danbooru Prompt Crafter (GPT)
I’ve updated the Danbooru Prompt Crafter with a larger Danbooru tag list for more accurate, detailed prompts.
It’s built to write, clean, and optimize prompts for models that use Danbooru-style tags.
What it does
Converts descriptions to Danbooru tags: paste a natural sentence, messy tag list, or even an image → get a clean, copy-ready prompt.
Cleans up messy prompts: reorders tags, removes duplicates, and replaces vague terms with precise, valid Danbooru tags.
Multiple modes:
Strict Mode → accurate, clean Danbooru tagging.
Creative Mode (default) → adds aesthetic polish for more visually appealing results.
Output types:
Realism (default) → for photorealistic images.
Anime/Art → for stylized or illustrated outputs.
Tag validation & correction: fixes misspelled/invalid tags automatically.
Flexible formatting: choose commas, underscores, or spaces.
Understands images: upload an image; it analyzes character, pose, fashion, background, and lighting to generate a Danbooru-style prompt.
Notes & Etiquette
This model can generate mature content. Use responsibly and follow platform rules.
No explicit deepfakes of real people. Don’t be that person. 🙃
Support & Early Access
If this model helps your work, consider supporting the lab.
Patreon gets you early builds, exclusive tools (like the Prompt Crafter GPT integrations), and behind-the-scenes experiments. Your support directly funds training and testing.
Description
All sample images are 100% LoRA-free
Version 2.0 is a smooth upgrade to the already-great V1.0. It follows Danbooru tags even better and fixes small quirks like the rare extra limb. Not a dramatic upgrade, just a nice polish to make your life a little bit easier.
FAQ
Comments (12)
The faces remind me of... a semi-realistic CyberPony? The checkpoint itself is superb, one of the best for semi-realism, despite the confusing name "Cyberanime":):)
By the way, the checkpoint works great for both semi-anime and semi-realism. Try Euler a SGM Uniform (Forge) for anime, and DPM++2M or 2M SDE for semi-realism. Excellent!
Yes, that’s definitely possible. I do have to say, though: I really find Illustrious nicer to work with than Pony SDXL. And the naming... yeah… It has to have a name, right? 😄How would you name it?
@Cyberdelia Haha, good question, bro. :) I think this particular checkpoint belongs in the "CyberAnime Catalyst" series. it clearly has some bold, experimental solutions that aren't exactly definitive. Just look at the tooltip errors. :) In any case, this checkpoint gets a huge thumbs up!
@AlexSadler tooltip errors? Please explain?
@AlexSadler Basically, this model is essentially the way Illustrious is at its core. It works with Danbooru tags - maybe you find that experimental, but that’s how it’s supposed to be.
@Cyberdelia Heh, I don't really understand how the AI works under the hood, maybe I'm imagining it. :) At least it's highly tolerant of spelling and syntax errors, which is good. That's more what I meant when I said "tooltip error." :) Plus, Google Translate might mess it up, haha.
@Cyberdelia What I was really trying to say is that you can find non-obvious options here using simple tags. So, the checkpoint is very creative. If that's what you intended, haha, it worked :)
Speaking of danboru, how critical is it to use _ instead of spaces? :) I've been struggling with this question for a while.
@AlexSadler That’s something you have to experiment with. It will work without strict Danbooru tags, but I notice that I get the exact desired result when using the correct tags.
@AlexSadler It is indeed a different kind of model compared to other models.
@Cyberdelia Ahah, so I finally spotted something. It's a different beast under the hood 😄
Whatever's inside, the vibe is already here, and I'm hooked. Keep it up! :)
kick ass checkpoint, hats off
Great checkpoint-- consistent, accurate, and creative-- not to mention the aesthetic itself which I really enjoy. All images I've posted below from this checkpoint use exactly the same generation settings; the only thing changed between them is the prompt. I really appreciate the lack of sameface in the model, too. :)
Great work-- much love and thanks to Cyberdelia. <3
Details
Files
Available On (2 platforms)
Same model published on other platforms. May have additional downloads or version variants.