Independent tools from 42.uk
Different models. Your creative direction.
Generate with cloud AI or keep your image workflow local.
Your generated images will appear here.
Choose where to generate, connect your model
and describe what you want to see.
This page keeps your latest 12 images in memory. Download images and settings before closing it. Local outputs also remain in your ComfyUI output folder.
About the workspace
42 MOMM Diffusion is an independent AI image workspace published by 42.uk. Connect different cloud models or a local ComfyUI server, create images from text, edit from a reference image, and export the prompt and settings. The interface is free; cloud provider usage and local hardware costs are separate.
Cloud connectors support OpenAI GPT Image, Google Gemini image models and Black Forest Labs FLUX. Local mode discovers compatible installed Stable Diffusion 1.5, SDXL and SDXL Turbo checkpoints. Switch models to compare the same idea across different backends.
Run ComfyUI on your computer, optionally through the third-party Promptus Manager, then connect its loopback address. In private local mode, this workspace sends prompts and reference images directly to that server, never through the 42.uk cloud connector.
Local workflows can combine diffusion generation with a learned upscaling model. ESRGAN-family upscalers provide a GAN-based enhancement stage; transformer-based and other learned upscalers use different architectures. Cloud models use the methods provided by their developers.
This is workflow composition: select a generation model, optionally edit an input, then apply an installed upscaler. It does not blend every model or method into a single neural network. Advanced architectures and custom node graphs remain available in your full ComfyUI workspace.
Start the ComfyUI backend in your local manager and check the address it reports. Enter that loopback address above and press Connect. A managed setup may use http://127.0.0.1:8288; standalone ComfyUI commonly uses port 8188. Allow the browser’s local-network permission when prompted. Your server must permit requests from https://42.uk. Install a compatible image checkpoint before connecting; this page does not install software or download models.
The app sends generation data directly between your browser and your own local ComfyUI server. It does not fall back to a cloud provider. The built-in graph uses local generation and image-processing nodes. Your local installation controls its own plugins and storage; ComfyUI saves outputs and can embed prompts in image metadata. Visiting 42.uk still requests this page from our web server.
No. 42 MOMM Diffusion is independently published by 42.uk. Promptus Manager is an optional third-party way to run the local ComfyUI backend. This integration does not imply ownership, sponsorship or endorsement by Promptus, ComfyUI or any cloud provider.
No. The built-in local workflow supports recognised SD 1.5, SDXL and SDXL Turbo checkpoints with compatible text encoders and decoders. FLUX checkpoints, audio models and other architectures need different graphs and are omitted from this local picker. Cloud FLUX is supported separately. GPU memory, model compatibility and provider account access affect which jobs can run.
Stop waiting ends this page’s wait and remaining batch jobs. A job already accepted by the local server or cloud provider may continue, and cloud charges may still apply. Check your backend before retrying. Download your images and settings before closing the page; the gallery is not a permanent archive.
Technical references: Promptus, ComfyUI, ESRGAN, OpenAI image API, Gemini image API, BFL image API.