PSA : to counteract slowness in SVI Pro use a model that already has a prebuilt LX2V LoRA
I renamed the model and forgot the original name, but I think it’s fp8, which already has a fast LoRA available, either from Civitai or from HF (Kijai).
I’ll upload the differences once I get home.
https://redd.it/1q2m5nl
@rStableDiffusion
I renamed the model and forgot the original name, but I think it’s fp8, which already has a fast LoRA available, either from Civitai or from HF (Kijai).
I’ll upload the differences once I get home.
https://redd.it/1q2m5nl
@rStableDiffusion
Reddit
From the StableDiffusion community on Reddit
Explore this post and more from the StableDiffusion community
I've created an SVI Pro workflow that can easily extended to generate longer videos using Subgraphs
https://redd.it/1q2s4bn
@rStableDiffusion
https://redd.it/1q2s4bn
@rStableDiffusion
Qwen Image 2512 - 3 Days Later Discussion.
I've been training and testing qwen image 2512 since Its come out.
Has anyone noticed
\- The flexibility has gotten worse
\- 3 arms, noticeably more body deformity
\- This overly sharpened texture, very noticeable in hair.
\- Bad at anime/styling
\- Using 2 or 3 LoRA's makes the quality quite bad
\- prompt adherence seems to get worse as you describe.
Seems this model was finetuned more towards photorealism.
Thoughts?
https://redd.it/1q2qe12
@rStableDiffusion
I've been training and testing qwen image 2512 since Its come out.
Has anyone noticed
\- The flexibility has gotten worse
\- 3 arms, noticeably more body deformity
\- This overly sharpened texture, very noticeable in hair.
\- Bad at anime/styling
\- Using 2 or 3 LoRA's makes the quality quite bad
\- prompt adherence seems to get worse as you describe.
Seems this model was finetuned more towards photorealism.
Thoughts?
https://redd.it/1q2qe12
@rStableDiffusion
Reddit
From the StableDiffusion community on Reddit
Explore this post and more from the StableDiffusion community
ComfyUI Tutorial: Major Update For Qwen Image 251
https://youtu.be/7tFEdLMEadc
https://redd.it/1q2x2n3
@rStableDiffusion
https://youtu.be/7tFEdLMEadc
https://redd.it/1q2x2n3
@rStableDiffusion
YouTube
ComfyUI Tutorial: Major Update For Qwen Image 2512 #comfyui #qwenimage #comfyuitutorial
On this tutorial, we will test out the new QWEN Image 2512 model, that was trained to enhace the photorealisme, natural details and textrendering. To test out the prompt adherance of the model various prompts was included for different usecases like portrait…
ComfyUI-GeminiWeb: Run NanoBanana Gen directly in Comfy without an API Key
https://redd.it/1q2w2e1
@rStableDiffusion
https://redd.it/1q2w2e1
@rStableDiffusion
Anything2Real 2601 Based on [Qwen Edit 2511]
# [RELEASE] New Version of Anything2Real LoRA - Transform Any Art Style to Photorealistic Images Based On Qwen Edit 2511
Hey Stable Diffusion community! 👋
I'm excited to share the new version of - **Anything2Real**, a specialized LoRA built on the powerful Qwen Edit 2511 (mmdit editing model) that transforms ANY art style into photorealistic images!
https://preview.redd.it/upvwo4n0g5bg1.png?width=2689&format=png&auto=webp&s=d40b86ee149e02bdf7bda22d457f676549a1fafa
https://preview.redd.it/dfz41lm0g5bg1.png?width=2653&format=png&auto=webp&s=47998475e881ec3396562263ec4c743b1df68e3b
https://preview.redd.it/ciil2mm0g5bg1.png?width=2325&format=png&auto=webp&s=30ec38a3bfa32cf5b1bf2c0d1ee12e19c4422ae0
https://preview.redd.it/kxtnekm0g5bg1.png?width=2638&format=png&auto=webp&s=c44648d148441b889f060491744094403563ca24
https://preview.redd.it/7ojhhmm0g5bg1.png?width=3272&format=png&auto=webp&s=d3c7df6fcf0f1c23b9c5cb79b0c410d3db9c57a4
https://preview.redd.it/38n0mlm0g5bg1.png?width=2516&format=png&auto=webp&s=1d176eb6626a11e71c98afa3e541565689b386a9
https://preview.redd.it/x39mqpm0g5bg1.png?width=2721&format=png&auto=webp&s=f435b79c1e8d2904fd8b478e12f4713fa580f3ef
# 🎯 What It Does
This LoRA is designed to convert illustrations, anime, cartoons, paintings, and other non-photorealistic images into convincing photographs while preserving the original composition and content.
# ⚙️ How to Use
* **Base Model:** Qwen Edit 2511 (mmdit editing model)
* **Recommended Strength:** 1(default)
* **Prompt Template:**
transform the image to realistic photograph. {detailed denoscription}
* Adding detailed denoscriptions helps the model better understand content and produces superior transformations (though it works even without detailed prompts!)
#
# 📌 Important Notes
* “realism” is inherently subjective, first modulate strength or switch base models rather than further increasing the LoRA weight.
* Should realism remain insufficient, blend with an additional photorealistic LoRA and adjust to taste.
* Your feedback and examples would be incredibly valuable for future improvements!
#
# Contact
Feel free to reach out via any of the following channels:
Twitter: @Lrzjason
Email: [lrzjason@gmail.com](mailto:lrzjason@gmail.com)
CivitAI: xiaozhijason
https://redd.it/1q2wnax
@rStableDiffusion
# [RELEASE] New Version of Anything2Real LoRA - Transform Any Art Style to Photorealistic Images Based On Qwen Edit 2511
Hey Stable Diffusion community! 👋
I'm excited to share the new version of - **Anything2Real**, a specialized LoRA built on the powerful Qwen Edit 2511 (mmdit editing model) that transforms ANY art style into photorealistic images!
https://preview.redd.it/upvwo4n0g5bg1.png?width=2689&format=png&auto=webp&s=d40b86ee149e02bdf7bda22d457f676549a1fafa
https://preview.redd.it/dfz41lm0g5bg1.png?width=2653&format=png&auto=webp&s=47998475e881ec3396562263ec4c743b1df68e3b
https://preview.redd.it/ciil2mm0g5bg1.png?width=2325&format=png&auto=webp&s=30ec38a3bfa32cf5b1bf2c0d1ee12e19c4422ae0
https://preview.redd.it/kxtnekm0g5bg1.png?width=2638&format=png&auto=webp&s=c44648d148441b889f060491744094403563ca24
https://preview.redd.it/7ojhhmm0g5bg1.png?width=3272&format=png&auto=webp&s=d3c7df6fcf0f1c23b9c5cb79b0c410d3db9c57a4
https://preview.redd.it/38n0mlm0g5bg1.png?width=2516&format=png&auto=webp&s=1d176eb6626a11e71c98afa3e541565689b386a9
https://preview.redd.it/x39mqpm0g5bg1.png?width=2721&format=png&auto=webp&s=f435b79c1e8d2904fd8b478e12f4713fa580f3ef
# 🎯 What It Does
This LoRA is designed to convert illustrations, anime, cartoons, paintings, and other non-photorealistic images into convincing photographs while preserving the original composition and content.
# ⚙️ How to Use
* **Base Model:** Qwen Edit 2511 (mmdit editing model)
* **Recommended Strength:** 1(default)
* **Prompt Template:**
transform the image to realistic photograph. {detailed denoscription}
* Adding detailed denoscriptions helps the model better understand content and produces superior transformations (though it works even without detailed prompts!)
#
# 📌 Important Notes
* “realism” is inherently subjective, first modulate strength or switch base models rather than further increasing the LoRA weight.
* Should realism remain insufficient, blend with an additional photorealistic LoRA and adjust to taste.
* Your feedback and examples would be incredibly valuable for future improvements!
#
# Contact
Feel free to reach out via any of the following channels:
Twitter: @Lrzjason
Email: [lrzjason@gmail.com](mailto:lrzjason@gmail.com)
CivitAI: xiaozhijason
https://redd.it/1q2wnax
@rStableDiffusion
I made 3 rtx 5090 available for image upscaling online. Enjoy!
you get up to 120s of gpu compute time daily ( 4 upscales to 4MPx with supir )
limit will probably increase in future as i add more gpus.
direct link is banned for whatever reason so i link a random subdomain:
https://232.image-upscaling.net
https://redd.it/1q39l84
@rStableDiffusion
you get up to 120s of gpu compute time daily ( 4 upscales to 4MPx with supir )
limit will probably increase in future as i add more gpus.
direct link is banned for whatever reason so i link a random subdomain:
https://232.image-upscaling.net
https://redd.it/1q39l84
@rStableDiffusion
image-upscaling.net
Free AI Image Upscaler | No Sign-Up, No Watermarks, Up to 16K
Upscale your images online up to 4x using the free AI Image Enhancer!
Flux 2 dev, tested with Lora Turbo and Pi-Flow node, Quality vs. Speed (8GB VRAM)
https://redd.it/1q39fm7
@rStableDiffusion
https://redd.it/1q39fm7
@rStableDiffusion
Reddit
From the StableDiffusion community on Reddit: Flux 2 dev, tested with Lora Turbo and Pi-Flow node, Quality vs. Speed (8GB VRAM)
Explore this post and more from the StableDiffusion community
ComfyUI Wan 2.2 SVI Pro: Perfect Long Video Workflow (No Color Shift)
https://www.youtube.com/watch?v=PJnTcVOqJCM
https://redd.it/1q3c7a5
@rStableDiffusion
https://www.youtube.com/watch?v=PJnTcVOqJCM
https://redd.it/1q3c7a5
@rStableDiffusion
YouTube
ComfyUI Wan 2.2 SVI Pro: Perfect Long Video Workflow (No Color Shift)
In this video, I show you how to generate continuous AI videos longer than 40 seconds with perfect character consistency. Most users struggle with faces melting or motion getting stuck after the first few seconds. I break down my exact "Manual SVI" method…
[Update] I added a Speed Sorter to my free local Metadata Viewer so you can cull thousands of AI images in minutes.
https://redd.it/1q34juf
@rStableDiffusion
https://redd.it/1q34juf
@rStableDiffusion
Reddit
From the StableDiffusion community on Reddit: [Update] I added a Speed Sorter to my free local Metadata Viewer so you can cull…
Explore this post and more from the StableDiffusion community
How do you create truly realistic facial expressions with z-image?
https://redd.it/1q36whm
@rStableDiffusion
https://redd.it/1q36whm
@rStableDiffusion
Reddit
From the StableDiffusion community on Reddit: How do you create truly realistic facial expressions with z-image?
Explore this post and more from the StableDiffusion community
Trellis 2 is already getting dethroned by other open source 3D generators in 2026
Today I saw two videos that show what 2026 will hold for 3D model generation.
A few days ago Ultrashape 1.0 released their model and can create much more detailed 3D geometry, then Trellis 2, without textures though, but an extra pass with the texture part of Trellis 2 might be doable.
https://github.com/PKU-YuanGroup/UltraShape-1.0
https://youtu.be/7kPNA86G\_GA?si=11\_vppK38I1XLqBz
Also the base models of Huyuan 3D and Sparc 3D, Lattice and FaithC, respectively are planed to release, together with other nice 3D goodness, already out or coming.
https://github.com/Zeqiang-Lai/LATTICE
https://github.com/Luo-Yihao/FaithC
https://youtu.be/1qn1zFpuZoc?si=siXIz1y3pv01qDZt
Also a new 3D multi part generator is also on the horizon with MoCa:
https://github.com/lizhiqi49/MoCA
Plus for auto rigging and text to 3d animations, here are some ComfyUi addons:
https://github.com/PozzettiAndrea/ComfyUI-UniRig
https://github.com/jtydhr88/ComfyUI-HY-Motion1
https://redd.it/1q3ijwo
@rStableDiffusion
Today I saw two videos that show what 2026 will hold for 3D model generation.
A few days ago Ultrashape 1.0 released their model and can create much more detailed 3D geometry, then Trellis 2, without textures though, but an extra pass with the texture part of Trellis 2 might be doable.
https://github.com/PKU-YuanGroup/UltraShape-1.0
https://youtu.be/7kPNA86G\_GA?si=11\_vppK38I1XLqBz
Also the base models of Huyuan 3D and Sparc 3D, Lattice and FaithC, respectively are planed to release, together with other nice 3D goodness, already out or coming.
https://github.com/Zeqiang-Lai/LATTICE
https://github.com/Luo-Yihao/FaithC
https://youtu.be/1qn1zFpuZoc?si=siXIz1y3pv01qDZt
Also a new 3D multi part generator is also on the horizon with MoCa:
https://github.com/lizhiqi49/MoCA
Plus for auto rigging and text to 3d animations, here are some ComfyUi addons:
https://github.com/PozzettiAndrea/ComfyUI-UniRig
https://github.com/jtydhr88/ComfyUI-HY-Motion1
https://redd.it/1q3ijwo
@rStableDiffusion
GitHub
GitHub - PKU-YuanGroup/UltraShape-1.0: High-Fidelity 3D Shape Generation via Scalable Geometric Refinement
High-Fidelity 3D Shape Generation via Scalable Geometric Refinement - PKU-YuanGroup/UltraShape-1.0