Turn your quick, roughly-sketched pen-and-ink black-line style cartoons into clean, scaleable images with smooth lines and consistent colours.
You can try a live demo right now.
This application is designed to process quick, roughly-sketched pen-and-ink black-line style cartoons into clean, scaleable images with smooth lines and consistent colours. It’s a workflow tool designed with beginners in mind, capable of bulk processing images, and can write processed images to a shared cloud storage location - perfect for busy school art classes.
The first thing you will see is an option to either take a photo or upload an image. If you are on a phone or tablet you can simply take a photo of the cartoon you want to use. On a laptop or desktop machine, a desktop USB visualiser is probably your best option.
Next, select if you want to palette-lock to a known set of colours. A wide range of palettes from pen, pencil and crayon sets used in UK schools are included. Any colours in the image, which for a scan or a photo will include subtle shades caused by lighting and natural variation, will be locked to their nearest, single, well-defined colour as given in the pen, pencil or crayon set you choose.
The application shows the total cost of the current session in the top bar. When you also supply an OpenRouter management key with -openrouter-management-key (or the OPENROUTER_MANAGEMENT_KEY environment variable), the remaining account balance is shown alongside it. Management keys are administrative-only: they can query your credits but cannot process images, so both keys are needed to see the balance. If the management key is missing or the query fails, the balance is simply hidden and only the session cost is shown.
The Models button in the top bar opens a searchable, selectable list of the OpenRouter models that can process an image and return a processed image, together with the cost of processing a single image with each one. When OpenRouter publishes an exact per-image price for a model it is used; otherwise the cost is estimated from the model’s published per-token rates assuming a typical 1024x1024 image plus a generated image output.
Select the “Vectorise” option if you want your output to be an SVG image, scaleable to fit any size and resolution of media - use your artwork in a book, on a t-shirt or a 96-sheet roadside billboard! This process is done by the application itself, so doesn’t need access to an AI processing engine (or even internet access) for this step.
You can try out Magooify on the live demo page.
You can run Magooify on your own system by simply downloading and running the executable for your platform from the bottom of this page. Simply run the executable, in desktop mode it should open your local web browser and display the user interface.
To use the AI image processing feature, you will need to provide an OpenRouter API key. Without one, the app still works: the prompt editing option is hidden and the captured image is never sent to an AI processing engine, Magooify just acts as a handy image-capture utility.
For more details, or if you are a system administrator wanting to run this application centrally for access by multiple users, see the documentation.
If you want to use, customise or extend this project - for use by yourself, other users in your school / workplace, or even as a service you charge money for - you are very welcome to. See the developer’s documentation for more details.
This project is distributed under a permissive Apache 2.0 license - you are allowed to freely use, modify, and distribute this project for both personal and commercial purposes without paying royalties.
64-bit Linux Desktop / Server
64-bit Windows Desktop
Intel-based Mac computers
Apple M1 / M2 / M3 / M4 Macs
Raspberry Pi 3/4/5 (64-bit OS)
Raspberry Pi 2/3/4 (32-bit OS)