SciFig

SciFig is an AI scientific figure generator that turns common research inputs—text prompts, hand-drawn sketches, reference images, research PDFs, and lab photos—into publication-ready scientific figures.

Start from what you already have: generate a figure in one of six publication styles (with Auto mode available), then refine it without rebuilding from scratch. SciFig keeps generated content editable, including on-text changes and region-based regeneration using natural-language prompts.

When you need full control, export figures as editable PPTX or layered SVG and fine-tune everything in the built-in canvas with undo/redo—so you can prepare journal-grade visuals without extra design roundtrips.

What inputs can I use to create figures in SciFig?

You can generate figures from text prompts, hand-drawn sketches, reference images, research PDFs, and lab photos.

Can I edit the figure after it’s generated?

Yes. SciFig keeps outputs editable so you can change text content, font, size, and position, and regenerate specific regions with a natural-language prompt.

What publication styles are available for generated figures?

SciFig provides six publication styles plus an Auto mode. Output styles include scientific, line, 3D, editorial, sketch, and watercolor.

How do exports work in SciFig?

You can export as editable PPTX or layered SVG, then continue refining details in the built-in canvas with undo/redo.

Is there a free way to start using SciFig?

New users can start free with credits on day one, including an option to get up to 400 free credits when joining through an invite.

Does SciFig use a default model for journal-style output?

SciFig includes GPT Image 2 in its Text-to-Figure flow, and you can switch models depending on your needs (e.g., for slides and posters).