Model
Trends
.ai
Model
Trends
.ai
the best open source ai models
Run Locally
DarkKlein9b_v2BFS_extracted_lora_r256.safetensors
1.23 GB · SafeTensor
Downloads are served by Civitai. Some creators ask for a Civitai login first.
AD
Install / Run on Graydient
Requires membership
<redcraft-exported-flux2>
About This Model
FLUX 2 / KLEIN · LoRAs
Created by
AiMetatron
on Civitai
by AiMetatron - RedCraft Exported LoRAs
Creator notes
RedCraft & DarkBeast H3|Z|Klein LoRAs Collect Submitted LoRA for & H3 weight 0.7-1 General prompt : Exploded View Diagram, For the target video, at 0.00 seconds into the target video, (from [Shot 1]) is fully referenced. integrated_multimodal_description: [Shot 1] Photoreal studio cinema, VFX already locked to : [SUBJECT] already whole at the center, already the same form as in the still. Setting and existing markings already as in the still. No new lettering. A studio camera already starts a slow 180-degree orbit around the center, small amplitude, slow speed, one continuous move, no cut. Through the first part of the orbit: already [SUBJECT] comes apart along its own structure, already [KEY PARTS] drift outward and stay readable, already each piece turns as it leaves. Through the middle: already the pieces hold apart, already the inner structure is clear. Through the last part: already the same pieces travel back along those paths and seat into one whole again, already the orbit finishes the 180. Hold on the reassembled subject. No captions, titles, subtitles, or watermarks beyond markings already in . overall_soundscape: Parts leaving, a short open hush, pieces seating back together, a soft studio hum, no speech. non_diegetic_music: none. 站内活动链接( on-site event )： 投稿作品收藏夹( Submitted LoRAs Collection )： All entries must be LoRAs trained on MiniMax H3. Train locally, or with our , which supports H3 today. All submissions must be SFW and must abide by 2,000,000 Buzz in prizes, and growing! Dark Beast Z8 SDA 🟥 ZITLoRA Edition 03/15/2026 DBZ 8 SDA 完整模型BF16： DBZ 8 SDA workflow 配套工作流链接： 通过引用 提出的新型 语义方向对齐（Semantic Directional Alignment, SDA）损失，本模型 merge 了 饼儿佬 训练的 LoKr 适配器，在完美保持 8 步基线模型极致锐度和快速推理速度的同时，恢复了 原始教师模型 70.2% 的组合多样性（LPIPS） Designed to rescue the "Diversity Collapse" problem in few-step distilled Flow Matching / Diffusion models. By applying novel Semantic Directional Alignment (SDA) Loss, this Model adapter recovers 70.2% of the original teacher model's compositional diversity (LPIPS) Dark Beast KLEIN 9b 🟦 V2.0 BFS 03/03/2026 DBK v2 BFS最佳换脸工作流，配套模型： DBK BFS workflow 配套工作流链接： This is the next-level face-swap specialized evolution of the Dark Beast lineage, built on the lightning-fast FLUX.2 Klein 9B accelerated model from Black Forest Labs. Engineered with targeted optimizations for face swapping practices, it integrates BFS (Best Face Swap) technology to completely eliminate the rigid, unnatural look that plagued earlier face replacements — delivering seamless, lifelike integrations with preserved identity, expression, and lighting. It also fully fixes the portrait reference issue from the previous DB BlitZ versions Special thanks to the scheme provider: for the powerful BFS foundation that powers this breakthrough.🟦 The author's In-site link: Important notes: This version is exclusively designed around the Klein 9B accelerated edition — no base model exists. Usage is identical to Black Forest Labs' official FLUX.2 Klein 9B accelerated release: ultra-low steps (e.g., 4-5), CFG=1 fixed, blazing inference speed on consumer hardware. In one sentence: Dark Beast's ferocious soul meets BFS (Best Face Swap) technology — more natural, and truly unstoppable! 🟦 for more infomation about BFS (Best Face Swap) : Alternatively, it can be directly applied to the entire Klein 9b/Qwen Edit base and Fine-tune models, through LoRA Adapter parameter injection. 2024 RedCraft | 红潮 CADS Q4 GGUF Enhancer-LoRA De-distill&CADS Commercial Materials Support online generation of ComfyUI WebUI matching： RedCraft | 红潮 CADS Commercial & Advertising Design System GGUF format Q4 quantization version for 1024 x 1024 inference better with This Pages ：RedCraft-Q4enhancer-LoRA The LoRA weight setting range is from 0.2 to 0.65 Inference parameters: 10-20 STEPS Euler / DPM++2M | beta / SGM_Uniform CFG 3-3.5 Real CFG needs to be set (ignore guidance, or set to 0)
5,536 downloads · 231 likes on Civitai
SHA256
0C9179E4DDB8E3D014DB504866C4DF138521C0808C0B5CD0781A982BE21006C3
ModelTrends.ai Model ID
#25189
Flag content
Example renders
Prompts and settings shared by the people who made these renders on Civitai. Pick a render to see how it was made.
Settings
Size
1024x1934
Prompt
这是一幅写实风格的电影感人像摄影，整体色调偏暖，带有复古胶片质感与柔和的黄昏光晕，营造出慵懒、性感且略带怀旧情绪的氛围。画面中有一位年轻女性（约20-30岁），她倚靠在一座古典石质建筑柱子旁，身体微微侧转，头部略微上扬，目光直视镜头，眼神迷离而富有挑逗性，嘴唇微张，嘴角自然下垂，呈现出一种慵懒又略带诱惑的姿态。她的肤色白皙，卷曲的浅棕色长发随意披散在肩头与背部，部分发丝被光线勾勒出轮廓。她身穿一件米白色的针织短款上衣（V领设计，袖子至肘部），露出平坦的腹部和清晰的腰线，下身搭配一条高腰、带有橙色与绿色花卉图案的A字长裙，裙摆因轻微动作而自然飘动，隐约可见腿部线条。她的左手搭在柱子边缘，右手自然垂落于身体一侧。背景为模糊的欧式古典建筑拱廊结构，光线从侧面打来，在人物身上形成明暗对比，突显了皮肤与衣物的质感。构图为中景特写，采用低角度略微仰拍，使人物显得更具存在感与视觉冲击力，画面主体集中在人物上半身及腰部以上区域，背景虚化处理以突出人物主体，整体光线柔和、暖色调为主，具有强烈的电影级光影表现力。 A Dark artistic logo in the upper center of the screen that reads'Dark Beast Z&KLEIN Exported LoRA Edition'
Settings
Size
1024x1934
Prompt
This photograph captures a young woman in a softly lit, indoor setting, which evokes a sense of naturalism and intimacy. The composition is centered on the subject, who gazes directly at the camera with a subtle, confident expression. Her long, wavy brown hair cascades over her shoulders, framing her face and adding a sense of softness and movement to the image. The use of light is particularly noteworthy, as it creates gentle, warm highlights and shadows on the woman's face, enhancing her natural beauty and the texture of her hair. The sunlight streaming in from the left side of the frame casts a delicate, almost painterly light on her skin, which is smooth and slightly flushed, adding a touch of warmth and realism. The color palette is predominantly soft and warm, with the white and blue floral pattern of her top contrasting gently against her skin and the neutral background. The floral pattern adds a touch of delicacy and femininity to the composition, while the woman's slightly parted lips and relaxed pose convey a sense of calm and confidence. In terms of artistic movement, this photograph could be associated with contemporary realism, with its emphasis on natural beauty, light, and everyday life. The texture of the woman's hair and the subtle play of light and shadow are reminiscent of classical portrait painting, but the photograph's modernity lies in its use of natural light and candid composition. Overall, the photograph is a nuanced exploration of light, texture, and natural beauty, combining elements of realism and contemporary portraiture to create a compelling and intimate image.
Settings
Size
1024x1934
Prompt
Ultra-detailed photoreal macro shot of a honeybee drinking from a droplet of sap on rough tree bark. The image freezes the moment mid-motion: the bee’s tiny articulated legs cling to the uneven wood grain, while translucent wings sit folded with visible veining illuminated by soft sunlight. Its fuzzy thorax and abdomen display individual hair fibers with natural directionality and tiny pollen grains trapped between them. The droplet of sap is perfectly clear, magnifying the texture beneath it, catching warm highlights, and distorting the background subtly through refraction. Bark texture appears rugged and cracked, with micro-shadows inside the crevices. Depth of field extremely shallow, isolating the bee in sharp focus while the background dissolves into creamy warm bokeh. Mood intimate, natural, and hyper-real, capturing fragile beauty and micro-world detail.
Settings
Sampler
_simple
Steps
20
Guidance
1
Size
832x1216
Prompt
realistic photograph of cyberpunk theme Hip-level Dutch angle, 24mm lens rotated 20 degrees clockwise, dynamic imbalance created through canted horizon and converging verticals A relaxed young Japanese woman , soft expression and bare shoulders, in an off-white knit crop top, high-waisted linen pants, and woven slides, mood: contented ease Posing playfully in front of a giant video screen in Shibuya 109, one hand on hip, the other flashing a peace sign, shoppers rushing past Giant digital billboards flashing anime ads and fashion campaigns above the Scramble Crossing, crowds moving like synchronized currents directional lighting with dramatic shadows, photorealistic quality, professional cinematography.
Settings
Size
1024x1934
Prompt
Photoreal 1960s candid street snapshot of two women waiting for a bus on a hot afternoon. One adjusts her beehive hairstyle while the other shields her eyes with a gloved hand. Their vintage dresses show small wrinkles and stitching imperfections, and one carries a floral-patterned box purse fading from sun exposure. The sidewalk is cracked, patched with tar, and littered with loose newspaper pages flapping in the breeze. A classic car rolls past in the background, chrome bumper reflecting distorted shapes from the street. The color palette washed in slightly muted Kodachrome tones: warm reds, pale yellows, and deep shadows. Mood timeless, grounded, elegantly imperfect.
Settings
Sampler
_simple
Steps
20
Guidance
1
Size
832x1216
Prompt
A dynamic, low-angle fashion photograph of a young woman with a messy, shaggy haircut and bold red lipstick, looking directly at the camera. She is crouching or leaning forward, wearing a chunky oversized white knit sweater, a white pleated mini-skirt, tall white socks, and prominent white chunky platform sneakers. The background is an abstract, angular studio set with dramatic, contrasting gel lighting intense blue and orange/yellow light creating a vibrant, energetic atmosphere and strong shadows. High-fashion editorial, hyper-realistic, sharp focus on the sneakers and face, cinematic color grading, cyberpunk undertones
Settings
Size
1024x1934
Prompt
A candid moment of a woman leaning against a subway pole, lost in thought as the train rumbles between stations. Harsh fluorescent lights overhead reveal natural skin texture, subtle under-eye shadows, and stray hairs across her forehead. Her fingers loosely grip a coffee cup, condensation collecting near the lid. Ads and graffiti fill the background in full detail as the car gently sways. Her expression is soft, introspective, completely unaware of any camera — pure urban realism.
Settings
Size
720x1360
Prompt
Photoreal 1960s candid street snapshot of two women waiting for a bus on a hot afternoon. One adjusts her beehive hairstyle while the other shields her eyes with a gloved hand. Their vintage dresses show small wrinkles and stitching imperfections, and one carries a floral-patterned box purse fading from sun exposure. The sidewalk is cracked, patched with tar, and littered with loose newspaper pages flapping in the breeze. A classic car rolls past in the background, chrome bumper reflecting distorted shapes from the street. The color palette washed in slightly muted Kodachrome tones: warm reds, pale yellows, and deep shadows. Mood timeless, grounded, elegantly imperfect.
AI Models for Image, Video & Text Generation
redcraft exported flux2 is a LoRA for the Flux 2 / Klein family listed on ModelTrends.ai, a read-only catalog of open source AI image, video and text models. Compare it with other Flux 2 / Klein models, check its heat score to see how it is trending, and open it on PirateDiffusion or BitVector to try it.
Browse by Model Family
Browse all models