Here is a simple routine for trying several layouts before settling on one. Open PaperFig, give it your starting material, and let it build the first version. In practice you upload a figure, a lab photo or a paragraph of your paper, add a short brief, and refine the draft that comes back. Text, sketches, photos, reference images and PDFs all work as a starting point. Once a draft exists, OCR turns its labels back into editable text. After that, it is aimed at the figures papers actually need: mechanisms, workflows and graphical abstracts. An existing figure can be fed back in and enhanced instead of started over. It is built for researchers who revise figures often. Typical jobs are manuscript figure drafts, graphical abstracts for submissions, workflow diagrams for grants, and presentation figures. Visit PaperFig.