new mecha flux2 — Model AI FLUX 2 / KLEIN

New Mecha style for Flux.2 Klein 9B! With the 9B version of Klein being more widely used that the 4B version, i did a quick attempt at training it. I can make it run on my poor setup using a , but boy it is slow. Anyway, here it is and the showcase pictures include some of the cursed stuff the model sometimes does for the sake of transparency regarding its capacities. Nonetheless, it works nicely for my rare needs of edit :D (Done using ai-toolkit, i stop the training at half the total steps, reworked a bit the last saved checkpoint using a n_average merge of previous epochs and restarted it before doing a final merge of epochs to get something more smooth). New Mecha style for Z-Image-Turbo V2! This time, i did a training using Z-Image De-Turbo, an undistilled version of ZIT. I also raised the LR to 0.0002 and made a 64 dim LoRA before reducing the rank to 32 and carefully merge it with the previous LoRA. I also removed the trigger word, the only constant remaining is the fact that it was training on "Digital illustrations" with an "anime-like" "semi-realistic style". This helped me achieve better results even if i raise the CFG to 1.2 (distilled models don't have a CFG and should be used with CFG = 1, aka no negative prompt). All showcase pictures are made with 9 steps, CFG 1.2, Euler/Beta (aka Smoothstep in stable-diffusion.cpp) using the Q4_K quant, the ablitered version of Qwen3-4B for the TE and the regular VAE, plus a flow shift at 3. They got an img2img pass at 0.5 denoise with the same parameters to smooth some details. Below a part of my ugly python wrapper script to parse multiple prompt ;-) Enjoy! 💜 New Mecha style for Flux.2 Klein 4B! This is a very first attempt at training a LoRA for Flux.2 Klein 4B! It was done using using the 4B-Base version, but all pictures are generated using the distilled version here. All showcase pictures are done using the and , first by doing a txt2img and then a img2img to sharpen the result a bit. This is a very first test and i still prefer ZIT, but the img edit capabilities are interesting and i wanted to have a version of the model that new my style (i'll probably try to train a better version later). Here is demo of the img edit. The source image was made using ZIT and my LoRA: And with this prompt and even a different resolution of 832x1216, Change the girl top from red to green and her hair from blonde to orange I got this picture 😁 (ok, it's a very basic test, but i'll try more stuff and at least, i am happy this run locally on my GTX 1060): New Mecha style for Z-Image-Turbo! This is a one shot test to allow me (and everyone of course) to test ZIT stuff with style. Works best with 8-10 steps. Trigger: Start your prompt with "NewMecha style illustration." followed by something (this can be altered) like "This is a highly detailed, digital illustration of a...". You can add "anime-like" if you wish for a flatter look or "CGI" if you want more 3D. This LoRA was trained using default config (since the on-site training is failing hard on ZIT) and a dataset of 250 pictures generated using New Mecha and captionned using . Since this is a dataset of Illustrious pictures, the overall result is a bit under what ZIT supposedly can render, but lowering the LoRA weight and adding more stuff is the goal (that's why on-site i put 0.8 as the max weight). The showcase was done by captionning my usual showcase pictures using JoyCaption too and the ugliest Gradio + Diffusers app ever (to avoid paying Yellow Buzz): import gradio as gr import torch from diffusers import ZImagePipeline from uuid import uuid4 from PIL.PngImagePlugin import PngInfo from datetime import datetime, timezone device = "cuda" dtype = torch.bfloat16 pipe = ZImagePipeline.from_pretrained( "Tongyi-MAI/Z-Image-Turbo", dtype=dtype, ).to(device) pipe.load_lora_weights( "./output/new_mecha_zit", weight_name="new_mecha_zit_000002750.safetensors", adapter_name="new_mecha", ) pipe.set_adapters(["new_mecha"], adapter_weights=[0.8]) def generate(prompt, seed, steps): generator = torch.Generator(device).manual_seed(seed) image = pipe( prompt=f"NewMecha style illustration. {prompt}", num_inference_steps=steps, guidance_scale=1, width=832, height=1216, generator=generator, ).images[0] im_path = f"./im-outputs/{uuid4()}.png" metadata = PngInfo() timestamp = datetime.now(timezone.utc).strftime("%Y-%m-%dT%H:%M:%S.%fZ") ressources = 'Civitai resources: [{"type":"checkpoint","modelVersionId":2442439,"modelName":"Z Image","modelVersionName":"Turbo"},{"type":"lora","weight":0.8,"modelVersionId":2507250,"modelName":"(ZIT) New Mecha style","modelVersionName":"v1.0"}], Civitai metadata: {}' metadata.add_text( "parameters", f"NewMecha style illustration. {prompt}\nNegative prompt: \nSteps: {steps}, Sampler: Undefined, CFG scale: 1, Seed: {seed}, Size: 832x1216, Clip skip: 2, Created Date: {timestamp}, {ressources}", ) image.save(im_path, pnginfo=metadata) return im_path demo = gr.Interface( fn=generate, inputs=[ gr.Textbox( label="prompt", value="This is a highly detailed, digital anime-style illustration of a young woman in a black dress.", lines=6, ), gr.Number(label="seed", value=42, minimum=1, precision=0), gr.Number(label="steps", value=8, minimum=1, precision=0), ], outputs=[gr.Image(height=608)], flagging_mode="never", clear_btn=None, ) demo.launch(server_name="0.0.0.0", server_port=8000) Enjoy and please give me any feedback you feel like regarding this LoRA (that's probably the only one i'll make for ZIT, at least until i can generate locally instead of on a costly GPU instance).

Kata pencetus

An anime-style digital illustration.

670 muat turun · 40 suka di Civitai

SHA256

DA507FF8B28BE59D74D65C5898B4D65D7FE3BB4A58DDF0EC847FA3A68F5375BA

ID Model ModelTrends.ai

#24500

Laporkan kandungan

Contoh render

Prompt dan tetapan yang dikongsi oleh orang yang membuat render ini di Civitai. Pilih render untuk melihat cara ia dibuat.

Dapatkan prompt ini

Render oleh a6bIp di Civitai

Laporkan

Tetapan

Langkah

Seed

777673855

Saiz

832x1216

Berat yang digunakan

0.7

Prompt

A cinematic retro-futuristic illustration of a lone bounty hunter seated in the open doorway of a hovering helicopter above a vast neon-lit metropolis at night. The city below glows with dense grids of magenta, cyan, and violet lights reflecting through drifting mist and rain. The hunter sits at the edge of the open door, preparing for a dangerous mission, holding a heavy futuristic revolver across her lap, its chamber faintly glowing with energy rounds. She has long white hair moving in the wind, sharp blue eyes, and worn leather gloves gripping the weapon with calm focus. Slightly ahead inside the cockpit, the back silhouette of the helicopter pilot is visible against glowing instrument panels. Far below, a massive corporate skyscraper rises above the surrounding skyline — the crime boss’s fortress and her target, subtle red police lights flicker through streets. The perspective is slightly off to the side with a Dutch angle, emphasizing tension and motion. Retro synthwave cyberpunk atmosphere, vibrant neon contrasts, cinematic lighting, dramatic storytelling, graphic novel style illustration. The image is both epic and dramatic. best quality, high resolution, hyper-detailed, cinematic, centered, very aesthetic, accurate, award-winning composition, soft grain, evocative, artistic, immersive storytelling. 4ft3rd4rk, aidmamj6.1, ArsMovieStill, 80s Fantasy Movie Still

Prompt negatif

low quality, worst quality, casual, boring

Render oleh Mrpopo714 di Civitai

Laporkan

Tetapan

Sampler

_simple

Langkah

Guidance

Saiz

832x1216

Prompt

An anime-style digital illustration. anime, high_budget_anime_movie_style, key_visual, ultra_detailed_background, 4k, 8k, cinematic_lighting, dawn_train_station_platform, painterly_environment, soft_cel_shading, adult_woman, imagine a lavish anime movie illustration of a completely different adult woman character in her early thirties waiting on an elevated train platform at dawn, composed like a quiet but emotionally charged key visual where she is rendered with refined, mature anime lineart and nuanced film-quality shading while the station, city, and sky are painted with expansive atmospheric depth; she is clearly an adult, around thirty-one, with a thoughtful, composed presence, average height and slender build, standing near the platform edge with a relaxed but purposeful posture, one hip slightly angled as if used to commuting, shoulders settled, body still but alert to the arriving train. Her face is elegant and introspective, oval-shaped with a soft jawline and gently defined cheekbones, cool porcelain skin warmed subtly by early morning light, faint shadows beneath the eyes suggesting an early start rather than fatigue; her anime-style eyes are cool gray-blue, narrow and intelligent, irises layered with soft gradients that reflect the pale sky and overhead wires, gaze focused down the tracks with quiet anticipation; makeup is understated and adult, muted taupe eyeshadow, fine eyeliner, natural lashes, lips in a soft dusty-rose tone with minimal shine. Her hair is distinctly different from before: short and modern, a clean chin-length bob in muted steel-blue with darker roots, slightly tousled by the morning breeze, strands catching soft rim light and lifting gently at the ends, framing her face in a calm, contemporary silhouette. Her outfit is urban, mature, and intentional, suited to a professional adult life: a long tailored coat in warm charcoal draped open, fabric rendered with heavy folds and subtle texture; beneath it, a fitted cream knit top tucked into high-waisted dark slacks with crisp seams; a slim leather belt and understated ankle boots complete the look; a structured work bag rests at her feet, one hand loosely holding the strap, the other adjusting the cuff of her coat as if grounding herself before the day begins; a simple watch peeks from beneath her sleeve, and small geometric earrings catch the light. She is actively doing something subtle but meaningful: shifting her weight as she steps closer to the safety line, glancing down the tracks as distant headlights glow, breath visible in the cool air, coat hem fluttering slightly as a breeze passes through the station. The environment is expansive and cinematic: the elevated platform stretches into the distance with steel beams, overhead power lines, and signage rendered in precise anime detail; benches, ticket machines, and warning stripes are painted with clean lines and gentle wear; below and beyond the platform, the city unfolds in layered depth, low-rise buildings, rooftops, and quiet streets softened by morning haze; the sky dominates the upper half of the composition, a luminous dawn gradient shifting from pale lavender and pink near the horizon into cool blue, thin clouds streaked with warm light; birds cross the sky in small silhouettes, and a distant train approaches, headlights glowing softly through the mist. Lighting is restrained but cinematic: early morning sunlight filters in at a low angle, casting long, soft shadows across the platform; cool ambient light fills the scene while warm highlights edge her coat, hair, and cheekbones; subtle reflections appear on metal railings and polished surfaces; the station feels calm and suspended in time. Anime shading is mature and controlled: gentle cel-shaded shadows on her face and body with smooth transitions, slightly deeper tones beneath the eyes and jaw for realism; hair shaded with clean planes and delicate highlights; clothing rendered with broader shadow shapes and ambient occlusion around folds and seams; the painterly background fades into atmospheric perspective as structures recede into mist. Composition places her slightly off-center, with platform lines and rails leading the eye toward the oncoming train and rising sun, creating a sense of forward motion and quiet resolve. Mood: reflective, poised, and quietly hopeful, like the opening or closing scene of a high-budget anime film about routine, choice, and the quiet power of showing up for a new day.

Render oleh Mrpopo714 di Civitai

Laporkan

Tetapan

Sampler

_simple

Langkah

Guidance

Saiz

832x1216

Prompt

An anime-style digital illustration. High-end anime movie style illustration of a cute young woman standing on the stone steps of an old forest shrine at dawn, character in clean anime lineart and shading, surrounded by lush, almost Ghibli-like painterly environment with magical realism; she is early 20s, light skin with soft blush, large green anime eyes, medium-length chestnut hair tied half-up with a small white ribbon, a few strands falling around her face, expression calm and a little dreamy; she wears a light beige hoodie over a white shirt, a soft moss-green skirt, simple brown boots, a small backpack on her shoulders; she stands mid-step, one foot slightly lifted, body turned slightly to look back over her shoulder, one hand resting on the stone rail, the other brushing her hair; the stone staircase climbs through a dense forest: tall cedar trees, dappled morning light filtering through leaves, moss-covered stone lanterns, faded red torii gates receding into the distance, everything rendered with rich hand-painted textures and subtle, soft edges; tiny glowing spirit-orbs float near the ground and around the lanterns, adding a subtle fantasy touch; rays of dawn sunlight cut through the trees in volumetric beams, soft warm light hitting her hair and shoulders, cool blue-green shadows under the trees, gentle mist near the ground; a small fox statue by the path is covered in moss, leaves flutter in midair, distant shrine bells implied by hanging ropes and paper charms; style is “movie key visual”: very detailed background, layered depth, subtle depth of field, while the girl has crisp outlines and carefully shaded anime-style coloring with his own soft gradient shading on skin, hair highlights painted with a few clean strokes; mood is peaceful, nostalgic, slightly magical, like a quiet scene before the story’s supernatural events begin; Tags: anime, anime_movie_style, forest_shrine, cute_girl, cinematic_background, painterly_forest, morning_light, volumetric_rays, soft_mist, key_visual, high_budget, ultra_detailed, 4k, 8k, aesthetic_scene, magical_realism, cinematic. , , ,

Render oleh a6bIp di Civitai

Laporkan

Tetapan

Sampler

DPM++ 2M_simple

Langkah

Guidance

Seed

770531712579021

Saiz

512x752

Clip skip

Prompt

A stylized digital painting depicting a slender yet muscular female knight with short white hair and piercing blue eyes, clad in blackened steel chainmail that glints beneath the sun. Her sleeveless breastplate is etched with deep sword marks, dents, and layered battle scars, faint healed cuts trace across her face and bare arms, telling stories of countless campaigns. She stands relaxed but authoritative on the right side of the composition, weight shifted to one leg, gauntleted hand resting near the hilt of her blade, chin slightly raised in quiet confidence. To her left, a disciplined line of young female recruits stands at attention, dressed in 17th-century musketeer-inspired uniforms: light blue cloth jerkins fitted neatly to their forms, polished brass buttons catching the sunlight, short cloaks marked with a bold cross, and slender swords secured at their waists. Their faces are youthful yet determined, gazes lifted toward their commander with admiration and fierce inspiration. The setting is a castle training ground enclosed by stone walls and banners fluttering in the breeze. Warm sunlight floods the courtyard, casting long, joyful highlights and soft shadows across cobblestones. The style is expressive and painterly — visible smeared paint, energetic brushstrokes, subtle dripping pigment, vibrant saturated colors, and a picturesque, artistic atmosphere that blends heroism with raw texture. Best quality, masterpiece, emotive, vibrant, artistic. ArsMJStyle, Etherial Gothic <lora:ModernFantasyPaintArt_f2Klein:0.7> <lora:EtherialGothicKlein9b:0.8> <lora:New_Mecha_Klein9B:0.7>

Prompt negatif

low quality, worst quality, bad quality, low resolution, casual, boring, dull, cliche, naive, 3d, 3d render, photo, cartoon, sketch, photograph, signature, logo censored, mosaic censoring, bar censor, 1girl, solo

Render oleh n_Arno di Civitai

Laporkan

Tetapan

Sampler

euler_a sgm_uniform

Langkah

Guidance

Seed

Saiz

832x1216

Prompt

a beautiful female angel with long, white, flowing hair, very pale skin, heavy makeup and a thick black lipstick. The focus is on her upper body, with her lower body out frame. She has large, white wings and a glowing halo above her head. Her eyes are closed, mouth open wide in ecstasy with her tongue slightly visible. She is lying flat on her stomach on a white bed with her upper body propped up, wearing a white negligee that barely covers her large, round breasts, with visible droplets of sweat on her skin. Her expression is one of orgasm. The background is dark with a starry night sky visible through an arched window. The lighting is soft and diffused, coming from above and slightly to the side, creating subtle highlights on her skin and hair. The camera viewpoint is eye-level, focusing on her face and breasts with a shallow depth of field that keeps the background slightly blurred. The composition uses the rule of thirds, with her face and chest positioned along the intersecting lines. The illustration is highly detailed and realistic in its rendering of textures and light. <lora:New_Mecha_Klein9B:1.0>

Render oleh a6bIp di Civitai

Laporkan

Tetapan

Sampler

capitan_flowmatch_advanced_sgm_uniform

Langkah

Guidance

Seed

31961449683874

Saiz

512x752

Clip skip

Prompt

Prompt negatif

Render oleh n_Arno di Civitai

Laporkan

Tetapan

Sampler

euler_a sgm_uniform

Langkah

Guidance

Seed

Saiz

832x1216

Prompt

a cute korean woman facing a wall with both hands on the surface while looking back toward the viewer. The wall is grey and covered in bold graffiti. She has fair skin and long black hair styled into high twintails, with circular-framed glasses resting on her nose and a bright, mischievous expression on her face. Her curvy figure is emphasized by a fitted white short-sleeve shirt and a very short, translucent, shimmering with iridescent colors, pleated skirt that catches the light and expose her ass as she shifts her weight onto one hip. She wears black and white striped thigh-high socks and carries a red, heart-shaped backpack slung casually over one shoulder. The lighting is vibrant and high-contrast, with soft daylight illuminating her from the front and subtle shadows adding depth. The camera angle is set at waist height with a slight upward tilt, enhancing confidence and attitude. Behind her, the wall features pink, dripping K-POP graffiti, partially framed by creeping green vines, creating a modern, playful, and provocative urban mood. <lora:New_Mecha_Klein9B:1.0>

Render oleh a6bIp di Civitai

Laporkan

Tetapan

Sampler

Euler

Langkah

Guidance

Seed

423054110

Saiz

832x1216

Berat yang digunakan

0.7

Prompt

A stylized digital painting depicting a slender yet muscular female knight with short white hair and piercing blue eyes, clad in blackened steel chainmail that glints beneath the sun. Her sleeveless breastplate is etched with deep sword marks, dents, and layered battle scars, faint healed cuts trace across her face and bare arms, telling stories of countless campaigns. She stands relaxed but authoritative on the right side of the composition, weight shifted to one leg, gauntleted hand resting near the hilt of her blade, chin slightly raised in quiet confidence. To her left, a disciplined line of young female recruits stands at attention, dressed in 17th-century musketeer-inspired uniforms: light blue cloth jerkins fitted neatly to their forms, polished brass buttons catching the sunlight, short cloaks marked with a bold cross, and slender swords secured at their waists. Their faces are youthful yet determined, gazes lifted toward their commander with admiration and fierce inspiration. The setting is a castle training ground enclosed by stone walls and banners fluttering in the breeze. Warm sunlight floods the courtyard, casting long, joyful highlights and soft shadows across cobblestones. The style is expressive and painterly visible smeared paint, energetic brushstrokes, subtle dripping pigment, vibrant saturated colors, and a picturesque, artistic atmosphere that blends heroism with raw texture. Best quality, masterpiece.

Model serupa

Model AI untuk penjanaan imej, video dan teks

new mecha flux2 ialah LoRA untuk keluarga Flux 2 / Klein yang disenaraikan di ModelTrends.ai, katalog baca sahaja untuk model AI imej, video dan teks sumber terbuka. Bandingkannya dengan model Flux 2 / Klein lain, semak skor habanya untuk melihat trendnya, dan bukanya di PirateDiffusion atau BitVector untuk mencubanya.

Lagi model Flux 2 / Klein Lagi LoRA Flux 2 / Klein anime sci-fi effects style new mecha

Layari mengikut keluarga model