Problem solved

Researchers across disciplines face the challenge of creating publication-quality scientific figures from their ideas, which often requires significant time and design expertise. SciFigureAI solves this by using generative AI to automatically produce figure drafts from natural language descriptions or rough sketches, dramatically accelerating the visual communication of research findings and reducing reliance on graphic design skills.

Use cases

• Generating graphical abstracts for research papers from a text summary of findings.
• Creating mechanism diagrams for biological processes (e.g., CAR-T cell therapy) by describing the steps in plain language.
• Designing protocol workflow figures for experimental methods (e.g., organoid culture) with numbered stages and lab iconography.
• Producing structure figures for materials science (e.g., crystal lattices) from text descriptions or sketches.
• Converting rough hand-drawn sketches into polished, high-quality figures for grant proposals or lab meetings.
• Iteratively editing and refining an existing figure draft by giving follow-up textual instructions to adjust composition or details.

Audience

• Researchers

Pros

• No design skills required; generates figures from simple text or sketches.
• Multiple input modes (text, sketch, image) cater to different starting points.
• Exports to vector (SVG) and presentation (PPTX) formats for easy integration into papers and slides.
• Iterative editing within the platform allows fast refinement without external tools.
• Provides scientific-domain inspiration examples to jump-start figure creation.

Cons

• Generated figures may need manual adjustments to meet journal-specific formatting or aesthetic standards.
• The AI's output styles are limited to those it has learned, potentially lacking versatility for highly specialized fields.
• Freemium model restricts higher-resolution exports or extended usage without purchasing credits.