CivArchive
    Preview 48794570
    Preview 48794576
    Preview 48794560
    Preview 48794572
    Preview 48794575
    Preview 48794574

    Anima Version is Available!!

    Nova Anime XL

    Nova Anime XL is Nova Anime: Anime/2.5D/3D checkpoint model for SDXL

    Pony version aims to get the similar look as both Nova Anime and Nova Domain

    All images on example were created with diffusers with custom png tags

    SeaArtAI

    OurtAI

    PixAI - IL v19.0

    PixAI - IL v18.0

    PixAI - IL v17.0

    PixAI - IL v16.0

    PixAI - IL v15.0

    PixAI - IL v14.0

    PixAI - IL v13.0

    PixAI - IL v12.5

    PixAI - IL v11.0

    PixAI - IL v10.0

    PixAI - IL v9.0

    PixAI - IL v8.0

    PixAI - IL v7.0

    PixAI - IL v6.0

    PixAI - IL v5.5

    PixAI - IL v5.0

    PixAI - IL v4.0 Happy Valentine

    PixAI - IL v3.0 Happy New Year

    PixAI - Illustrious v1.0

    PixAI - Illustrious v2.0

    PixAI - IL v2.5 Merry Christmas

    PixAI - Pony v1-v2

    PixAI - Pony v3

    PixAI - Pony v4

    PixAI - Pony v5

    PixAI - Pony v6

    PixAI - v7.0

    Rules

    You cannot use the generated images for commercial use if it's not edited (or just turning it to black and white)
    You can share images without any restriction if you don't monetize it

    Advertising this model to outside is always welcome

    Recommend Settings

    Sampler: Euler a

    Steps: 20~30

    Clip Skip: 1-2

    Denoising Strength: 0.65 - 0.8
    (Pony)

    CFG Scale: 5~7

    Prompt: score_9, score_8_up, score_7_up, score_6_up, score_5_up, score_4_up, source_anime, BREAK

    Negative Prompts: score_4, score_5, 3d, jpeg artifacts, username, watermark, signature, normal quality, worst quality, large head, low quality, text, error, missing fingers, extra digits, fewer digits, bad eye

    (Illustrious)

    CFG Scale: 4~6
    Prompt: masterpiece, best quality, amazing quality, 4k, very aesthetic, high resolution, ultra-detailed, absurdres, newest, scenery, {Prompt}, BREAK, depth of field, volumetric lighting

    Negative Prompts: modern, recent, old, oldest, cartoon, graphic, text, painting, crayon, graphite, abstract, glitch, deformed, mutated, ugly, disfigured, long body, lowres, bad anatomy, bad hands, missing fingers, extra digits, fewer digits, cropped, very displeasing, (worst quality, bad quality:1.2), bad anatomy, sketch, jpeg artifacts, signature, watermark, username, signature, simple background, conjoined,bad ai-generated

    Description

    Applied Noob 1.0 structure

    FAQ

    Comments (44)

    besporyadkin966Jan 1, 2025· 4 reactions
    CivitAI

    oh hell na brah. I just started trying out noob, and now you give me this?! lol

    Noxy_Jan 1, 2025· 2 reactions
    CivitAI

    Oh boy, V3, this one is promising but the more I use these newer versions, the more I realized the recommended settings don't work anymore

    I have been trying to use this with any setting I feel I haven't used or ones that are recommended in other models but it still didn't work out

    I hope the recommended settings in this model description gets more in depth like scheduler type because even if I used the displayed images as well as some here still didn't look good

    602539Jan 1, 2025

    Just ignore everything and use masterpiece, best quality as quality tags and lowres, worst quality, low quality, bad anatomy, bad hands, 4koma, comic, greyscale, censored, jpeg artifacts, as negatives on almost all models.

    I usually like to add potted plant, text, patreon, signature, text, watermark, twitter username to negs as well.

    Zap9Jan 1, 2025

    The recommended settings are working great for me in ComfyUI.

    Noxy_Jan 2, 2025

    @ABSDE That does sound too simple though, I'll try it out and if it doesn't work I might just switch to another platform instead of ReForge

    Alright, tested for a while and it turns out, having a simple prompt gets higher quality than complex ones (or ones requiring a lot of specifics like wanting a demon girl in a cyberpunk), created several girls in simple clothing without anything else is fine but the moment I try to make my pfp it just doesn't output the expected quality IF I left the quality prompts like the recommended ones

    Once I trimmed it, it returns a decent output and a half cry of what I would get, I saw another comment saying this got prompt adherence of Flux and that is bad for me as I tried Flux before and I failed to even get one single good custom character of mine from it

    Same goes for negative, trimming it made it look better....

    I am probably giving up on the recent models and return to V1 for a while

    Noxy_Jan 2, 2025

    @Zap9 That's what I have been wondering, is it worth switching to ComfyUI? I have been just using ReForge because I didn't want to redo my entire setup again..

    Zap9Jan 2, 2025

    @Noxy_ For me, it's definitely worth it since I use custom nodes a lot. The flexibility for custom workflows is so useful.

    Noxy_Jan 2, 2025

    @Zap9 I might consider that, as ReForge can be really confusing at times with zero guide on their github

    Zap9Jan 1, 2025· 2 reactions
    CivitAI

    What scheduler is best for the IL models?

    Crody
    Author
    Jan 1, 2025

    For me, it was Euler A

    Zap9Jan 1, 2025

    @Crody Yeah, I've been using that with the normal scheduler, but I'm asking about the scheduler to use with Euler A, like normal, karras, etc. I think I remember the simple scheduler being recommended for base NoobAI XL.

    Crody
    Author
    Jan 2, 2025

    @Zap9 It depends on what you want but for me, Normal is the best

    ShaionaJan 2, 2025

    Euler a and simple works best for me

    gausssidorov928Jan 1, 2025· 5 reactions
    CivitAI

    10/10 nice!!!

    hboxgames132Jan 3, 2025· 1 reaction
    CivitAI

    Happy new year sir

    VexLabsJan 3, 2025· 2 reactions
    CivitAI

    Great work

    can you make another one but niji journey style and make looks like gokuryo style and another niji styles

    Crody
    Author
    Jan 3, 2025

    2.5D style with realistic background and high contrast?

    LongRodVonHugenDongJan 4, 2025· 7 reactions
    CivitAI

    Excellent checkpoint!

    Noxy_Jan 4, 2025· 5 reactions
    CivitAI

    Holy shit, after hours of experimenting, I figured out and fixed why my images were blurry

    I just had to edit the prompt template to include volumetric lighting and a dutch angle, I cannot believe that's all it took...

    Well, I did had to fiddle with ReForge again but thank god it worked, I'll be taking my word and say V3 is finally good again! (Shout out to the fellow people that commented on my V3 rant for helping me steer the prompting)

    Noxy_Jan 4, 2025

    Dear lord, Hires Fix doesn't seem to work with this model anymore

    miao1973Jan 5, 2025· 2 reactions
    CivitAI

    May I ask a question?

    Could you allow me to upload pony v1.0~v7.0 on Tensor?
    I don't monetize anything including support for the account.

    I will be sure to link to the original and include the rules in the description.

    Crody
    Author
    Jan 5, 2025

    Sure

    miao1973Jan 5, 2025

    @Crody Thank you.

    By the way, although it is not monetization of the model itself, is it allowed to monetize accounts like Buzz?

    Crody
    Author
    Jan 5, 2025

    @miao1973 No

    miao1973Jan 5, 2025

    @Crody Thank you for your reply.
    I understand that monetization of anything other than the model itself is not permitted.
    For this model, I follow that rule.

    Hikarun_16Jan 5, 2025· 3 reactions
    CivitAI

    It is so hard for me to generate an original male character with full recommend prompt without making the character having toned body but girl's face/head, I don't know if anyone else having this problem or it's just me, overall I'm just looking for a solution, feel free to give me a solution if you have, otherwise, if it is an inherent problem from the model itself, I hope it gets fixed so that if you put 1boy or the likes it would immediately assume it means at least a generic anime mc, instead of something girly

    A1322Jan 19, 2025· 1 reaction

    This is my experience, as well, and I think it's an issue with Nova models across the board. It is possible, but it's basically luck, the deck is heavily stacked against you, and if you want something other than your stock anime protagonist then godspeed, my friend.

    It's very difficult to make stuff other than pinups and lewds with anime-focused models, in general. I always want to use AI for workshopping character concepts and it's very frustrating when half your cast is male. (The female half is a breeze, though!)

    raffyffyFeb 14, 2025

    yes same,but i tried animagine and its almost the same but with a little more luck.

    raffyffyFeb 14, 2025

    update : i think its a little better on the latest version 4.0 of nova, for example i was trying 1boy,inuyasha,holding hands with 1girl hatsune miku and ive gotten fairly good results

    raffyffyFeb 14, 2025

    update : nah its the same HAHAHAHA only inuyasha was good as per my testing

    hbbydrg353Jan 12, 2025· 2 reactions
    CivitAI

    In the recommended settings, what does 'denoising strength: 0.65 - 0.8' mean? Is it specifically for the i2i sampler?

    Crody
    Author
    Jan 13, 2025· 1 reaction

    i2i and hires

    TetsuooJan 15, 2025· 2 reactions
    CivitAI

    Nova 1.0: Not bad at all, but Better Days Illustrious gives very similar results with more details... In other words this is basically the same as WAI NSFW Illustrious. It's hard work to compare all the models haha ok let's try the others now x)

    gausssidorov928Jan 15, 2025· 4 reactions
    CivitAI

    Я не знаю какую магию применил автор, но рисует эиа модель на отлично, 100\10

    AwesomeesJan 18, 2025· 3 reactions
    CivitAI

    one word: AWESOME!

    zxbbck730Jan 20, 2025· 6 reactions
    CivitAI

    Can the generated pictures be used to make videos and upload them to YouTube to make money?

    Crody
    Author
    Jan 20, 2025

    If the video is about advertising my model, then yes

    Other than that, if it's just the slideshow thing then you need to edit the image to follow the rules on description

    zxbbck730Jan 20, 2025

    Can I add LoRa to change the style and make money?

    Crody
    Author
    Jan 20, 2025

    @zxbbck730 The thing is that the image should be untraceable so adding LoRAs would do the same thing

    Also, LoRAs also have their own rules so I wouldn't recommend it

    zxbbck730Jan 20, 2025

    Got it. Thank you~

    green_angerJan 28, 2025

    @Crody Out of curiosity, what makes an image traceable? It cannot be simply png chunks.

    Crody
    Author
    Jan 28, 2025

    @green_anger Guess the prompt by looking or tag fetching using image to tag

    I know it's not likely to happen but still there is possibility that anyone can create the image or at least similar one

    For example, this image made with AsumaXL (credit goes to Qnuk) https://civitai.com/images/53648954

    Can be guessed the prompt something like: 1girl, solo, blonde hair, long hair, flowing hair, orange eyes, looking at viewer, smile, hand in hair, floating hair, blue medieval frilled dress, green field, blue sky, upper body, wide shot

    It's not exact same but at least traceable

    zxbbck730Feb 3, 2025

    @green_anger After adding LORA, the style cannot be changed.