| Table of Contents |
|---|
Reference Image
civitai image 113243826 by 6vidit9 using Z Image on ConfyUI
PROMPT Reference Prompt A high-contrast monochromatic fine art portrait captures a close-up of a delicate, feminine hand holding an irregularly shaped fragment of a shattered mirror. The hand, positioned in the lower-left to mid-left foreground, displays visible knuckles and fingernails with a dark polish, gripping the jagged, textured edge of the mirror shard with the thumb and index finger. Within the highly reflective surface of the mirror fragment, a melancholic reflection of a woman's face is rendered in sharp detail.
...
The lighting is dramatic and directional, creating strong chiaroscuro effects that accentuate the contours of the hand and the features of the reflected face. Bright highlights illuminate the left side of the reflected face and the sharp edges of the mirror, contrasting deeply with the shadowed areas. The overall image exhibits a shallow depth of field, rendering the background into a softly blurred, dark, out-of-focus bokeh, suggesting an indistinct or natural setting, which serves to isolate the central subject and enhance the psychological intensity of the composition. The entire scene evokes themes of introspection, vulnerability, and fragmentation.
Other metadata cfgScale:1.5 steps:15 sampler:Euler
Open Source models
Human created prompt: Reflection of woman face, close-up from lips to eyes, in a a piece of broken mirror which she holds in her hand. Her fingernails are with a dark polish. Almost monochrome. Blurry dark background behind the mirror. | Chat GPT enhanced: Close-up reflection of a woman’s face captured within a jagged piece of broken mirror she holds delicately between her fingers. Focus sharply on the area from her lips to her eyes, revealing subtle skin texture and a calm, introspective expression. Her fingernails are coated in dark, glossy polish, adding contrast to the near-monochrome palette. Soft, cinematic lighting creates gentle highlights on the mirror’s edges while the background falls into a deep, blurred darkness. The composition feels intimate, editorial, and emotionally rich — a high-end fine-art portrait photograph with exquisite detail and mood. | QWEN 3 VL 8B Caption from ChtGPT generated image In this hauntingly beautiful black-and-white portrait, a woman’s face is revealed through a fragmented, geometric mirror held delicately in her hands. The mirror, shaped like a sharp, angular shard, frames only a portion of her visage — her piercing eyes, defined brows, and lips — while the rest of her face dissolves into shadow, emphasizing the theme of partial revelation. Her fingers, adorned with glossy black nail polish, cradle the mirror with quiet intensity, suggesting both control and vulnerability. The stark contrast of light and dark, combined with the minimalist composition against a void-like background, evokes a sense of introspection, mystery, and emotional depth. The image captures a moment of quiet contemplation — as if she is peering into a reflection of herself, or perhaps confronting a hidden truth. The monochrome palette amplifies the drama, turning the scene into a timeless meditation on identity, perception, and the fractured nature of self. |
Tempest by Vlad XL base v0.1 (SDXL) | ||
|---|---|---|
Parameters: Steps: 20| Size: 1024x1024| Seed: 818764411| CFG scale: 3.4| App: SD.Next| Version: a9b0ab2| Pipeline: StableDiffusionXLPipeline| Operations: txt2img| Model: tempestByVlad_baseV01| Model hash: 8bfad17222 Time: 31.61s | total 44.52 pipeline 29.94 preview 10.71 decode 1.64 callback 1.53 prompt 0.35 gc 0.32 | GPU 10598 MB 8% | RAM 14.67 GB 12% |
Parameters: Steps: 20| Size: 1024x1024| Seed: 2649270740| CFG scale: 3.4| App: SD.Next| Version: a9b0ab2| Pipeline: StableDiffusionXLPipeline| Operations: txt2img| Model: tempestByVlad_baseV01| Model hash: 8bfad17222 Time: 31.27s | total 38.61 pipeline 29.64 preview 4.92 callback 1.63 decode 1.59 prompt 0.50 gc 0.29 | GPU 10598 MB 8% | RAM 14.6 GB 12% |
Parameters: Steps: 20| Size: 1024x1024| Seed: 210598755| CFG scale: 3.4| App: SD.Next| Version: a9b0ab2| Pipeline: StableDiffusionXLPipeline| Operations: txt2img| Model: tempestByVlad_baseV01| Model hash: 8bfad17222 Time: 31.98s | total 41.34 pipeline 30.34 preview 6.67 callback 1.65 decode 1.60 prompt 0.75 gc 0.29 | GPU 10598 MB 8% | RAM 14.57 GB 12% |
Stable Diffusion 3.5 Large | ||
Parameters: Steps: 28| Size: 1024x1024| Seed: 131020695| CFG scale: 3.5| App: SD.Next| Version: a9b0ab2| Pipeline: StableDiffusion3Pipeline| Operations: txt2img| Model: stable-diffusion-3.5-large Time: 3m 19.53s | total 212.52 pipeline 196.80 preview 8.98 decode 2.71 callback 1.92 prompt 1.79 gc 0.29 | GPU 30478 MB 24% | RAM 37.14 GB 30% | Parameters: Steps: 28| Size: 1024x1024| Seed: 3571390739| CFG scale: 3.5| App: SD.Next| Version: a9b0ab2| Pipeline: StableDiffusion3Pipeline| Operations: txt2img| Model: stable-diffusion-3.5-large Time: 3m 25.83s | total 215.58 pipeline 203.02 preview 5.99 decode 2.79 callback 1.92 prompt 1.54 gc 0.29 | GPU 30478 MB 24% | RAM 37.47 GB 30% |
Parameters: Steps: 28| Size: 1024x1024| Seed: 628653312| CFG scale: 3.5| App: SD.Next| Version: a9b0ab2| Pipeline: StableDiffusion3Pipeline| Operations: txt2img| Model: stable-diffusion-3.5-large Time: 3m 28.12s | total 214.52 pipeline 203.89 decode 4.21 prompt 2.46 callback 1.92 preview 1.70 gc 0.31 | GPU 30478 MB 24% | RAM 37.29 GB 30% |
Flux.1 Dev | ||
Parameters: Steps: 50| Size: 1024x1024| Seed: 3875921176| CFG scale: 3.5| App: SD.Next| Version: a9b0ab2| Pipeline: FluxPipeline| Operations: txt2img| Model: FLUX.1-dev Time: 8m 5.61s | total 499.25 pipeline 482.93 preview 9.06 decode 2.65 callback 2.63 prompt 0.83 te 0.83 gc 0.29 | GPU 35838 MB 28% | RAM 41.65 GB 34% |
Parameters: Steps: 50| Size: 1024x1024| Seed: 810223809| CFG scale: 3.5| App: SD.Next| Version: a9b0ab2| Pipeline: FluxPipeline| Operations: txt2img| Model: FLUX.1-dev Time: 8m 14.39s | total 521.92 pipeline 491.69 preview 22.87 callback 2.68 decode 2.68 prompt 0.84 te 0.83 gc 0.29 | GPU 35838 MB 28% | RAM 41.87 GB 34% |
Parameters: Steps: 50| Size: 1024x1024| Seed: 752350128| CFG scale: 3.5| App: SD.Next| Version: a9b0ab2| Pipeline: FluxPipeline| Operations: txt2img| Model: FLUX.1-dev Time: 8m 13.92s | total 521.92 pipeline 491.20 preview 23.31 callback 2.72 decode 2.69 prompt 0.83 te 0.83 gc 0.31 | GPU 35822 MB 28% | RAM 42.99 GB 35% |
FLUX.1-Krea-dev | ||
Parameters: Steps: 50| Size: 1024x1024| Seed: 3049274953| CFG scale: 6| App: SD.Next| Version: a9b0ab2| Pipeline: FluxPipeline| Operations: txt2img| Model: FLUX.1-Krea-dev Time: 7m 36.59s | total 477.71 pipeline 454.16 preview 16.81 decode 2.41 callback 2.40 prompt 0.80 te 0.80 gc 0.31 | GPU 35872 MB 28% | RAM 42.65 GB 35% |
Parameters: Steps: 50| Size: 1024x1024| Seed: 1040662746| CFG scale: 6| App: SD.Next| Version: a9b0ab2| Pipeline: FluxPipeline| Operations: txt2img| Model: FLUX.1-Krea-dev Time: 7m 34.23s | total 478.41 pipeline 451.79 preview 19.86 callback 2.49 decode 2.42 prompt 0.77 te 0.76 gc 0.29 | GPU 35872 MB 28% | RAM 42.85 GB 35% | |
Cosmos Predict 2 2B | ||
 Parameters: Steps: 50| Size: 512x512| Seed: 4007970578| CFG scale: 6| App: SD.Next| Version: a9b0ab2| Pipeline: Cosmos2TextToImagePipeline| Operations: txt2img| Model: Cosmos-Predict2-2B-Text2Image | Parameters: Steps: 50| Size: 512x512| Seed: 1644724748| CFG scale: 6| App: SD.Next| Version: a9b0ab2| Pipeline: Cosmos2TextToImagePipeline| Operations: txt2img| Model: Cosmos-Predict2-2B-Text2Image |
Parameters: Steps: 50| Size: 512x512| Seed: 1922479893| CFG scale: 6| App: SD.Next| Version: a9b0ab2| Pipeline: Cosmos2TextToImagePipeline| Operations: txt2img| Model: Cosmos-Predict2-2B-Text2Image Time: 1m 9.99s | total 87.73 pipeline 69.98 callback 14.94 te 2.23 vae 0.58 | GPU 14638 MB 12% | RAM 20.66 GB 17% |
Cosmos Predict 2 14B | ||
Parameters: Steps: 50| Size: 1024x1024| Seed: 232364475| CFG scale: 6| App: SD.Next| Version: a9b0ab2| Pipeline: Cosmos2TextToImagePipeline| Operations: txt2img| Model: Cosmos-Predict2-14B-Text2Image Time: 22m 44.72s | total 1378.76 pipeline 1364.69 callback 10.42 te 2.52 vae 1.09 | GPU 42206 MB 33% | RAM 49.12 GB 40% | Parameters: Steps: 50| Size: 1024x1024| Seed: 1168702943| CFG scale: 6| App: SD.Next| Version: a9b0ab2| Pipeline: Cosmos2TextToImagePipeline| Operations: txt2img| Model: Cosmos-Predict2-14B-Text2Image Time: 22m 44.65s | total 1377.66 pipeline 1364.62 callback 10.42 te 2.42 | GPU 42206 MB 33% | RAM 49.19 GB 40% |
Parameters: Steps: 50| Size: 1024x1024| Seed: 4241165633| CFG scale: 6| App: SD.Next| Version: a9b0ab2| Pipeline: Cosmos2TextToImagePipeline| Operations: txt2img| Model: Cosmos-Predict2-14B-Text2Image Time: 22m 44.42s | total 1377.37 pipeline 1364.40 callback 10.35 te 2.44 | GPU 42206 MB 33% | RAM 49.19 GB 40% |
FLUX.2-dev | ||
Parameters: Steps: 28| Size: 1024x1024| Seed: 42| CFG scale: 4| App: SD.Next| Version: a9b0ab2| Pipeline: Flux2Pipeline| Operations: txt2img| Model: FLUX.2-dev |
Parameters: Steps: 28| Size: 1024x1024| Seed: 42| CFG scale: 4| App: SD.Next| Version: a9b0ab2| Pipeline: Flux2Pipeline| Operations: txt2img| Model: FLUX.2-dev |
Parameters: Steps: 50| Size: 1024x1024| Seed: 4227405778| CFG scale: 4| App: SD.Next| Version: a9b0ab2| Pipeline: Flux2Pipeline| Operations: txt2img| Model: FLUX.2-dev |
Z Image Turbo (9 steps) | ||
Parameters: Steps: 9| Size: 1024x1024| Seed: 42| CFG scale: 1.0| App: SD.Next| Version: a9b0ab2| Pipeline: ZImagePipeline| Operations: txt2img| Model: Z-Image-Turbo |
Parameters: Steps: 9| Size: 1024x1024| Seed: 42| CFG scale: 1.0| App: SD.Next| Version: a9b0ab2| Pipeline: ZImagePipeline| Operations: txt2img| Model: Z-Image-Turbo |
Parameters: Steps: 9| Size: 1024x1024| Seed: 42| CFG scale: 1.0| App: SD.Next| Version: a9b0ab2| Pipeline: ZImagePipeline| Operations: txt2img| Model: Z-Image-Turbo |
Z Image Turbo (12 steps) cherry picked | ||
Parameters: Steps: 12| Size: 1024x1024| Seed: 2496951322| CFG scale: 1.6| CFG true: 1.6| App: SD.Next| Version: a9b0ab2| Pipeline: ZImagePipeline| Operations: txt2img| Model: Z-Image-Turbo |
Parameters: Steps: 12| Size: 1024x1024| Seed: 2179728094| CFG scale: 1.4| CFG true: 2| App: SD.Next| Version: a9b0ab2| Pipeline: ZImagePipeline| Operations: txt2img| Model: Z-Image-Turbo |
Parameters: Steps: 12| Size: 1024x1024| Seed: 3761786925| CFG scale: 1.0| CFG true: 2| App: SD.Next| Version: a9b0ab2| Pipeline: ZImagePipeline| Operations: txt2img| Model: Z-Image-Turbo Time: 1m 40.05s | total 118.45 pipeline 98.30 preview 16.17 decode 1.73 callback 1.24 te 0.59 gc 0.40 | GPU 23646 MB 19% | RAM 29.03 GB 24% |
Qwen Image Distill Full (15 steps) | ||
Parameters: Steps: 15| Size: 1024x1024| Seed: 1661613733| CFG scale: 4| CFG true: 4| App: SD.Next| Version: a9b0ab2| Pipeline: QwenImagePipeline| Operations: txt2img| Model: Qwen-Image-Distill-Full |
Parameters: Steps: 15| Size: 1024x1024| Seed: 1996751832| CFG scale: 4| CFG true: 4| App: SD.Next| Version: a9b0ab2| Pipeline: QwenImagePipeline| Operations: txt2img| Model: Qwen-Image-Distill-Full |
Parameters: Steps: 15| Size: 1024x1024| Seed: 857374360| CFG scale: 4| CFG true: 4| App: SD.Next| Version: a9b0ab2| Pipeline: QwenImagePipeline| Operations: txt2img| Model: Qwen-Image-Distill-Full |
Qwen Image | ||
Parameters: Steps: 50| Size: 1024x1024| Seed: 1661613733| CFG scale: 4| CFG true: 4| App: SD.Next| Version: a9b0ab2| Pipeline: QwenImagePipeline| Operations: txt2img| Model: Qwen-Image 285H Time: 12m 54.32s | total 828.84 pipeline 774.29 preview 50.09 callback 2.28 te 1.18 vae 0.97 | GPU 60242 MB 48% | RAM 75.59 GB 61% Parameters: Steps: 50| Size: 1024x1024| Seed: 3990977625| CFG scale: 4| CFG true: 4| App: SD.Next| Version: a9b0ab2| Pipeline: QwenImagePipeline| Operations: txt2img| Model: Qwen-Image Time: 14m 6.53s | total 904.23 pipeline 846.49 preview 53.04 callback 2.38 te 1.40 vae 0.87 | GPU 60392 MB 48% | RAM 68.1 GB 55% | Parameters: Steps: 50| Size: 1024x1024| Seed: 105680236| CFG scale: 4| CFG true: 4| App: SD.Next| Version: a9b0ab2| Pipeline: QwenImagePipeline| Operations: txt2img| Model: Qwen-Image 285H Time: 13m 0.10s | total 822.74 pipeline 780.07 preview 38.99 callback 2.28 te 1.15 | GPU 60242 MB 48% | RAM 75.63 GB 61% Parameters: Steps: 50| Size: 1024x1024| Seed: 561946448| CFG scale: 4| CFG true: 4| App: SD.Next| Version: a9b0ab2| Pipeline: QwenImagePipeline| Operations: txt2img| Model: Qwen-Image Time: 14m 12.59s | total 901.56 pipeline 852.56 preview 45.03 callback 2.38 te 1.33 | GPU 60396 MB 48% | RAM 68.12 GB 55% |
Parameters: Steps: 50| Size: 1024x1024| Seed: 2699211064| CFG scale: 4| CFG true: 4| App: SD.Next| Version: a9b0ab2| Pipeline: QwenImagePipeline| Operations: txt2img| Model: Qwen-Image 285H Time: 13m 4.96s | total 826.91 pipeline 784.93 preview 38.22 callback 2.30 te 1.19 | GPU 60242 MB 48% | RAM 75.66 GB 61% Parameters: Steps: 50| Size: 1024x1024| Seed: 4218994297| CFG scale: 4| CFG true: 4| App: SD.Next| Version: a9b0ab2| Pipeline: QwenImagePipeline| Operations: txt2img| Model: Qwen-Image |
HiDream-I1-Fast | ||
Parameters: Steps: 16| Size: 1024x1024| Seed: 718435405| CFG scale: 1.5| App: SD.Next| Version: a9b0ab2| Pipeline: HiDreamImagePipeline| Operations: txt2img| Model: HiDream-I1-Fast Time: 5m 32.06s | total 342.43 pipeline 329.51 prompt 4.14 te 4.14 decode 2.53 vae 1.32 preview 0.46 gc 0.31 | GPU 63406 MB 50% | RAM 75.25 GB 61% | Parameters: Steps: 16| Size: 1024x1024| Seed: 2832466329| CFG scale: 1.5| App: SD.Next| Version: a9b0ab2| Pipeline: HiDreamImagePipeline| Operations: txt2img| Model: HiDream-I1-Fast Time: 6m 11.89s | total 387.98 pipeline 369.63 preview 6.23 prompt 4.03 te 4.03 decode 2.24 vae 1.50 gc 0.29 | GPU 63470 MB 50% | RAM 74.69 GB 61% |
Parameters: Steps: 16| Size: 1024x1024| Seed: 1097437983| CFG scale: 1.5| App: SD.Next| Version: a9b0ab2| Pipeline: HiDreamImagePipeline| Operations: txt2img| Model: HiDream-I1-Fast Time: 5m 51.23s | total 414.81 pipeline 348.56 te 25.97 prompt 25.96 preview 9.96 decode 2.64 vae 1.38 gc 0.32 | GPU 63342 MB 50% | RAM 74.44 GB 60% |
| Hunyuan Image 2.1 Distilled | ||
Parameters: Steps: 8| Size: 2048x2048| Seed: 3054762250| CFG scale: 3.25| App: SD.Next| Version: a9b0ab2| Pipeline: HunyuanImagePipeline| Operations: txt2img| Model: HunyuanImage-2.1-Distilled-Diffusers Time: 3m 48.29s | total 231.58 pipeline 228.21 vae 1.62 te 1.14 callback 0.52 | GPU 58134 MB 46% | RAM 72.56 GB 59% |
Parameters: Steps: 8| Size: 2048x2048| Seed: 1692669621| CFG scale: 3.25| App: SD.Next| Version: a9b0ab2| Pipeline: HunyuanImagePipeline| Operations: txt2img| Model: HunyuanImage-2.1-Distilled-Diffusers Time: 2m 47.46s | total 168.50 pipeline 167.38 callback 0.51 te 0.44 | GPU 58134 MB 46% | RAM 72.44 GB 59% |
Parameters: Steps: 8| Size: 2048x2048| Seed: 126093362| CFG scale: 3.25| App: SD.Next| Version: a9b0ab2| Pipeline: HunyuanImagePipeline| Operations: txt2img| Model: HunyuanImage-2.1-Distilled-Diffusers Time: 2m 47.45s | total 168.46 pipeline 167.37 callback 0.51 te 0.41 | GPU 58134 MB 46% | RAM 72.48 GB 59% |
| WAN 2.2 14B combined | ||
Parameters: Steps: 28| Size: 1024x1024| Seed: 567707335| CFG scale: 4| App: SD.Next| Version: e0cadf0| Pipeline: WanPipeline| Operations: txt2img| Model: Wan2.2-T2V-A14B-Diffusers |
Parameters: Steps: 28| Size: 1024x1024| Seed: 2832466329| CFG scale: 4| App: SD.Next| Version: e0cadf0| Pipeline: WanPipeline| Operations: txt2img| Model: Wan2.2-T2V-A14B-Diffusers |
Parameters: Steps: 28| Size: 1024x1024| Seed: 211603714| CFG scale: 4| App: SD.Next| Version: e0cadf0| Pipeline: WanPipeline| Operations: txt2img| Model: Wan2.2-T2V-A14B-Diffusers |
Non open-source
Human created prompt: Reflection of woman face, close-up from lips to eyes, in a a piece of broken mirror which she holds in her hand. Her fingernails are with a dark polish. Almost monochrome. Blurry dark background behind the mirror. | Chat GPT enhanced: Close-up reflection of a woman’s face captured within a jagged piece of broken mirror she holds delicately between her fingers. Focus sharply on the area from her lips to her eyes, revealing subtle skin texture and a calm, introspective expression. Her fingernails are coated in dark, glossy polish, adding contrast to the near-monochrome palette. Soft, cinematic lighting creates gentle highlights on the mirror’s edges while the background falls into a deep, blurred darkness. The composition feels intimate, editorial, and emotionally rich — a high-end fine-art portrait photograph with exquisite detail and mood. | Reverse prompt from previous ChatGPT image by QWEN 3 VL 8B In this hauntingly beautiful black-and-white portrait, a woman’s face is revealed through a fragmented, geometric mirror held delicately in her hands. The mirror, shaped like a sharp, angular shard, frames only a portion of her visage — her piercing eyes, defined brows, and lips — while the rest of her face dissolves into shadow, emphasizing the theme of partial revelation. Her fingers, adorned with glossy black nail polish, cradle the mirror with quiet intensity, suggesting both control and vulnerability. The stark contrast of light and dark, combined with the minimalist composition against a void-like background, evokes a sense of introspection, mystery, and emotional depth. The image captures a moment of quiet contemplation — as if she is peering into a reflection of herself, or perhaps confronting a hidden truth. The monochrome palette amplifies the drama, turning the scene into a timeless meditation on identity, perception, and the fractured nature of self. |
ChatGPT - https://chatgpt.com/ | ||
|---|---|---|
|
|
|
Google Gemini https://gemini.google.com/ | ||
|
|
|
xAI Grok | ||
|
|
|
|
|
|
SD.Next System Info
HW: GMKTec NucBox EVO-T1 (128GB/8TB)
| Code Block |
|---|
Sat Dec 13 11:11:22 2025 app: sdnext.git updated: 2025-12-11 hash: e0cadf015 url: https://github.com/vladmandic/sdnext/tree/master
arch: x86_64 cpu: x86_64 system: Linux release: 6.17.0-8-generic
python: 3.12.3 PyTorch 2.9.1+xpu
device: Intel(R) Graphics [0x7d51] (1) ipex:
ram: free:121.85 used:1.22 total:123.07
xformers: diffusers: 0.36.0.dev0 transformers: 4.57.3
active: xpu dtype: torch.bfloat16 vae: torch.bfloat16 unet: torch.bfloat16
refiner: none vae: none te: none unet: none |