create-image
Public
A native LM Studio plugin that generates images using Stable Diffusion (via stable-diffusion.cpp).
.safetensors model (SD 1.5, SDXL, etc.)Search for create-image in LM Studio's Integrations panel.
After installation, click on create-image in the Integrations panel to configure:
| Setting | Description |
|---|---|
| sd-cli Path | Full path to your sd-cli executable |
| Model Path | Path to your .safetensors model |
| Output Directory | Where to save generated images |
| Image Width/Height | Generation resolution (multiples of 64) |
| Sampling Steps | More = better quality, slower |
| CFG Scale | 1-2 for DMD2/LCM, 7-12 for standard |
| Sampling Method | LCM, Euler, DPM++, etc. |
| Preview Scale % | Size of inline preview (10-100%) |
| Enable Upscaling | Toggle ESRGAN upscaling |
| Upscale Model Path |
Simply ask the model to generate an image:
"Generate an image of a sunset over mountains"
The model will call the generate_image tool and display the result inline.
sips for preview image resizing (no external dependencies)This plugin would not be possible without the following amazing projects:
stable-diffusion.cpp by - A blazing-fast C++ implementation of Stable Diffusion that powers the image generation in this plugin. Thank you for making local SD inference so accessible!
MIT
A native LM Studio plugin that generates images using Stable Diffusion (via stable-diffusion.cpp).
.safetensors model (SD 1.5, SDXL, etc.)Search for create-image in LM Studio's Integrations panel.
After installation, click on create-image in the Integrations panel to configure:
| Setting | Description |
|---|---|
| sd-cli Path | Full path to your sd-cli executable |
| Model Path | Path to your .safetensors model |
| Output Directory | Where to save generated images |
| Image Width/Height | Generation resolution (multiples of 64) |
| Sampling Steps | More = better quality, slower |
| CFG Scale | 1-2 for DMD2/LCM, 7-12 for standard |
| Sampling Method | LCM, Euler, DPM++, etc. |
| Preview Scale % | Size of inline preview (10-100%) |
| Enable Upscaling | Toggle ESRGAN upscaling |
| Upscale Model Path |
Simply ask the model to generate an image:
"Generate an image of a sunset over mountains"
The model will call the generate_image tool and display the result inline.
sips for preview image resizing (no external dependencies)This plugin would not be possible without the following amazing projects:
stable-diffusion.cpp by - A blazing-fast C++ implementation of Stable Diffusion that powers the image generation in this plugin. Thank you for making local SD inference so accessible!
MIT
| Path to ESRGAN model |
Stable Diffusion XL (SDXL) by Stability AI - The revolutionary text-to-image model that enables high-quality image generation. Special thanks to the research team for advancing open-source AI art.
LM Studio by LM Studio - The incredible local LLM platform that makes running AI models locally accessible to everyone. Thank you for the excellent plugin SDK and documentation!
cd create-image
npm install
lms push
| Path to ESRGAN model |
Stable Diffusion XL (SDXL) by Stability AI - The revolutionary text-to-image model that enables high-quality image generation. Special thanks to the research team for advancing open-source AI art.
LM Studio by LM Studio - The incredible local LLM platform that makes running AI models locally accessible to everyone. Thank you for the excellent plugin SDK and documentation!
cd create-image
npm install
lms push