Illustration

Flux Jib / L10N-Blank Lora

SD3.5 Large SIMPLE IMG2IMG

Still under development. Stay tuned. Generates rubbish for now... 🐟

Tạo tranh màu nước theo ảnh người thật (watercolor paintings from real photos).

Prompt:If you are having problems with pics coming out weird:change the prompt in that specific part - make the prompt more clear(Ex: instead of "his fingers wrapping around the sword" use "he holds the sword") don't use one super large sentences to describe one thing - break it into pieces (Ex: instead of "a beautiful woman with red long hair wears a green old cape and sits in a medieval castle" use "a beautiful woman with red long hair, wearing a green cape, sitting in a medieval castle")don't break the flow of the prompt - write it like a sequence(Ex: instead of "a beautiful woman, wearing a green cape, sitting in a medieval castle, with red long hair" use "a beautiful woman with red long hair, wearing a green cape, sitting in a medieval castle")the "abstract prompt" (image to prompt) has a weird natural language that messes up almost every LoRa I tried - correct its language and order of descriptionKeywords: (for this to work at its best, it's good to use this type of words)illustration, medieval, fantasy, vintage, scary, hauntedwizard, man, woman, couple, monster, creature, ghost, vampirenight, moon, forest, tree, plants, stones, ruins, castle, overgrown, skull, sword, armor, horse, graveyard, castle, manor, haunted house, trees, bats, grass, field, ruins, abandoned, candles, candle-holder, scared, running, dark, skeletonYou can describe:the appearance: long/short hair, curly/straight/wavy hair, blonde/black/red hair, beardthe clothes: robe, cloak, dress, armor, corset, cape, vintage, sheer dress, night gown, underwear, bodysuit, lingerieand the state/emotion: scared, running, walking, screaming, closed eyesEffects:light particles, dust particles, texturemagic, sparkle, shine, fog

Enter a prompt and see it appear in Halloween effect !

Upload Origin Image + Face Swap => Generate___________▪️ Web: stablediffusion.vn▪️ Job: hungdiffusion.com▪️ Mail: sdvn@stablediffusion.vn

This app is enhanced with creepy subject and color effects suitable for 3Dillustration objects!!! Try a few and see which one you like... Please post your results so I can see how they look to you. The output will be an Image and Text Description of the generated prompt.

Just upload your picture and let the magic happen.FLUX Model

* This is a simple tool for SD3.5.- Only one LoRA setting.- You can use simple additional optional effects with "Style Effect ♥".

This AI tool streamlines the input parameters to their most fundamental form for casual users inexperienced with AI, optimizing the settings to maximize the value of your credits. This tool was made for entertainment.Instructions:Select a Generation Seed: You don't need to alter the remaining seed values for your initial image generation.Choose an anime reference to guide your prompt to that style; this is useful if your prompt is straightforward.Enter your preferred prompt in the tidytag format (the way we prompt in PonyV6) in this field. Natural language is not advised at all.In the 'face detailer' area, type a prompt to customize the facial aspect.Choose a 'Resolution': In PonyV6, resolution is essential to achieving the result you want; various outputs perform better at particular resolutions.Input Prompt tips and tricks:Use the 'Variance seed' and 'ESND' to get variations of specific outputs, remember keeping the 'Generation (seed)' static.Make sure to have a correct tokenization employing the best tags (view checkpoint model and LoRAs to know more). Danbooru tags tends to move the output into NSFW territory.The prompt in the 'face detailer' field won't affect your input prompt, it won't contribute for its tokenization, this prompt is independent of your input prompt.Syntax like 'WAS A1111', 'AND', 'BREAK' are invalid, please do not try it,ADetailer (facedetailer): the tool has been limited to recognize up to 5 faces at the same time.MIscellaneous📘 What is the 75 prompt limit in stable diffusion?🦄 Tokenization in Ponyv6The way tokens work is that they are batched in groups of 75. If you have <=75 tokens, the diffusion process has only one batch of tokens. If you have 78 tokens, you will have one batch of 75 tokens and one batch of 3 tokens, and the attention of the diffusion process will be divided in half. This will give you less of your first 75 tokens and may overemphasize the 3 prompt tokens in the second batch. The more batches of tokens, the less attention on each batch.🦄 Pony Diffusion XL Recognized Characters List🦄 Danbooru Tagging Visualization for PonyXL/AutismMix⚙️ Preloaded ItemsYou don't need to provide score tags like score_9, score_8_up, score_7_up, and source_anime because they are already proloaded.Although they are not included, rating tags like rating_safe, rating_questionable, and rating_explicit can be added.Although they are not included, source tags like source_furry and source_pony can be added.Negative prompt: not accessible since it has already been preloaded.Version: 1.0✉️ Discord Thread: Soon

SD 3.5 Turbo 4 Steps / 4步快速出圖

Tạo Chibi Sticker từ ảnh của bạn !Hãy tải ảnh lên và cùng xem Ai tạo Chibi Sticker dễ thương như thế nào nhé !^^

This is an anime style illustration drawing tool for SD3.5L. Equipped with 3 LoRa. Please replace LoRa as you like.