One home for every number behind your work - earnings broken out by source, real analytics, and bulk fee & access management. Open to every creator now.

Meet the Creator Studio: Your Numbers, Finally

We're a Day-0 launch partner for MiniMax H3, a next-gen open-weights multimodal video model - and it's already in the generator. No waitlist. Open weights land in a few days.

MiniMax H3 is live: open-weights video, generate now

The people who generate great images drive a huge part of Civitai - and they don't have great ways to earn yet. We've got some ideas, and we want yours. Read the proposal and drop your thoughts in the comments.

Proposal: How should image creators earn?

Set your own prices, sell access to your models, run your own shop, and see everything you earn in the new Creator Studio. Here's the 2026 Creator Program, in full.

The 2026 Creator Program

Design and sell custom cosmetics, list your models, and run a real storefront right on your profile. Creator Shops are here - Creator Program members now, everyone soon.

Open Your Own Creator Shop

Krea 2 LoRA training is 50% off, and we're running a Krea 2 Training Contest with 1.8 million Buzz in prizes across four LoRA categories.

⚡ Krea 2 Training Contest + 50% Off Training

Eureka from Eureka Seven (Psalms Of Planets Eureka Seven) trained for Pony XLV6 - A really weird version. I removed the alternate hairs as they were causing issues and instead I threw in basically any Manga/Doujin image I could find of Eureka. I then minimal tagged each image (10 tags per image) and then let Civitai train it for 7300 steps (If you remove the batch size it's considerably more steps)V5 - Tried to improve existing content, added two hair variations, added her underwear which appears in some media and tried adding her wingsV4 - Military outfit and flightsuit added (I do not guarantee this one will work. The additions of these two outfits in multiple tests was having effects on the model as a whole. This Epoch I have posted is the best of a bad bunch and I still prefer V3 moreV3 - General improvements<s>V2 contains an experimental 3 in one LoRA containing her Eureka Seven, Astral Ocean and Hi Evolution 3 Eureka appearances </s>(((Do not use, this 3-1 idea didn't work)))V1 contains the original Eureka

Removed alternate hairstyles as they were causing issues. I added images from the manga and from doujins to the dataset, this seems to have somewhat helped with her hairclip and collarI edited a Koikatsu Eureka model to give it her military outfit, bodysuit and underwear and made images using that to try expand the dataset. I did try a full Koikatsu dataset but the results were abysmal as the 3D style of that game seems to cause major issues so I binned that ideaGiven this thing have 7300 steps in it, which removing the batch size of 5, is well into the tens of thousands of steps it might well be overtrained but even in its overtrained state everything seems to be appearing to some extent as it should be which is the first time that's happened in a resultTry using it at 0.9, if not flexible try 0.8 but at 0.8 her sleeves were changing colour and at 1.0 one of the test images almost looked to be a copy of one in the dataset so this LoRA may be right on the edge of overtrained

This is HEAVILY EXPERIMENTAL. I've been working on this model for days now and this result has been the best so far. I've previously thought that my other models were bad but then people have generated okay images with them so I'm happy to post this one to see how it turns out but I'll probably make a V6 to fix any issues but not right now as I want to make something not Eureka. I'm drainedAdded (and hopefully working)<ul><li>Long hair from her Pocket Full of Rainbows appearance</li><li>Alternative hair (Her hair from the original series 3rd act)</li><li>Wings (Can't guarantee anything on these, I've never done winged characters and I had to fight the model on what it thought her wings were as one time it made her a dragon)</li><li>Underwear</li><li>Added a misc folder with a lot of images where Eureka appears to help it learn her character based on my experience training my FFIII LoRA's. They're mostly a bunch NSFW stuff because nude images are the best way to separate characters from outfits to avoid outfits merging into the body</li></ul>Changed<ul><li>Altered dataset for her Main Outfit, Military Outfit, Bodysuit (Don't use the hairclip, hair ornament tags or collar tags with the bodysuit, it goes a bit weird)</li><li>Altered dataset to try and focus down her hairclip to make the model learn it</li><li>Changed E7E activation tag to EUREKE7, partly because there's now 4 models floating around using this same E7E tag and partly because it dawned on me that a three letter activation tag may not be a great idea</li></ul>

V5 (Experimental)

<ul><li>Added military uniform (It's a recolour version of her normal outfit from the Hi Evolution special episode which also has some different boots)</li></ul><ul><li>Added flight suit (Not going to be perfect. I could only find 10 images of this and half of them are really bad quality)</li></ul><ul><li>Adjusted some tagging and purged the dataset of images which seemed useless</li></ul><ul><li>As ever the usual hairclip quirks remain so inpaint or ADetailer them</li></ul><ul><li>I wouldn't recommend generating the flight suit with her collar as the collar doesn't appear in it</li><li>Can't say if this version will be better or worse. Never intended to add any more outfits to Eureka as she was already a complicated character as is</li></ul>

<ul><li>Added better quality images</li></ul><ul><li>Removed some images which may reduce quality</li></ul><ul><li>Added "manga" images</li><li>Added images of Talho in Eurekas outfit to help model understand the outfit as a separate concept to Eurekas appearance</li></ul><ul><li>Hairclip/Hair Ornament seems to be somewhat improved but inpainting still may be needed</li><li>Thigh strap seems to be improved</li><li>Purple eye issue persists but may be a natural issue to Pony XL so use ad-detail or adjust "purple eye" weight</li><li>Her boots seem slightly improved</li><li>LoRA wasn't trained on a unique trigger but will generate using the characteristics</li></ul>

This version probably doesn't work. By all means try it but I'll probably just train the three series versions of Eureka separately to avoid the contamination issues(Added)Eureka - Eureka Seven Astral OceanEureka - Eureka Seven HiEvolution(Retrained)Eureka - Eureka Seven

V2 - "It just works"

Trained on 100 images, originally meant for the bin but decided to upload it as I found it is capable of producing some good images if prompted enough

Eureka [Pony XL] - Eureka Seven