1
Upload your sketch
Snap a photo or scan — napkin, whiteboard, or paper.
Collapse sidebar
Generator works best on desktop
Sign up now — your account and credits will be ready when you switch to a larger screen.
SKETCH-TO-FIGURE
From whiteboard sketch
to clean scientific figure.
Upload a hand-drawn sketch and SciFig renders it into a publication-ready figure — with real labels, clean lines, and your chosen style.
Figure prompt
Describe the figure you want from this sketch
Action / Details (e.g., Double strand break, detailed molecular view)
How it works
Sketch to scientific figure, in 3 steps
2
Describe what it becomes
Tell AI the subject, style, and any missing details.
3
AI renders your style
SciFig renders your sketch to figure with clean lines and real labels — no tracing required.
Sketch-to-Figure FAQ
Common questions about converting hand-drawn sketches into publication-ready figures.
1.
From napkin to publication-grade figure, in minutes.
- Paper, Whiteboard, Napkin: Any hand-drawn sketch works. Photograph it, upload, and SciFig sketch to figure converts it into a publication-ready scientific figure.
- No Tablet, No Illustrator: Zero drawing tablet. Zero Pen-tool-in-Illustrator learning curve. If you can sketch stick figures, you can use SciFig.
- Labeled, Layered, Editable: Not a traced scan — a clean figure with correct labels, aligned arrows, and journal-style typography. Every element can be individually edited.
2.
Yes. We interpret intent, not line quality.
- Messy Lines Are Fine: Crooked circles, shaky arrows, misaligned boxes — SciFig reads the structure you meant, not the strokes you drew.
- Stick-Figure Level Works: Researchers tell us they spent 10 hours learning Illustrator before discovering SciFig handles 3-minute bus-stop sketches just as well.
- Scientific Context Aware: Draw a rough signaling cascade — SciFig infers which blobs are kinases, which arrows are activation vs inhibition, and labels accordingly.
3.
Tracing keeps the mess. SciFig sketch to figure rebuilds clean.
- Rebuild, Don't Trace: Illustrator's Image Trace preserves your wobbly lines. SciFig generates new, clean vector shapes with consistent weights and journal-grade typography.
- 10 Hours → 3 Minutes: Researchers told us they spent days learning Illustrator before finding SciFig. Same output, 100× less time investment.
- Organized Layers, Not Flat SVG: Output comes with named, grouped layers — not one flat layer of spaghetti paths.
4.
Yes. Two ways — canvas or conversation.
- Click-to-Edit Canvas: Open the result in Vector Canvas to adjust text, colors, and layout directly. For element-level vector edits — moving individual strokes, re-styling arrows, recoloring shapes independently — run Vectorize to convert the figure into a layered SVG where every shape is a separately selectable object.
- Multi-Round AI Refinement: Type "make the PAM site bigger and change the arrow from green to red" — SciFig updates the figure without you touching anything.
5.
Anything with a clear spatial story.
- Ideal Candidates: Mechanism diagrams, experimental workflows, signaling pathways, cell structures, equipment setups, flowcharts — anything where "which shape connects to which" is the point.
- Distinctness Beats Neatness: Three obviously-different blobs beat three identical crisp circles. Make each component visually distinct, even if crude.
- Add Labels by Hand: Scribble "Kinase A" next to one blob and "PAM" next to another. SciFig reads handwritten labels and renders them as clean text.
6.
Yes. Drawing + description is our secret weapon.
- Both Inputs Together: Upload the sketch and add "this is a CRISPR delivery mechanism, add PAM site labels" — SciFig sketch to figure uses both to produce a far more accurate figure than either alone.
- Sketch Gives Structure, Text Gives Meaning: Your drawing handles layout. Your words handle terminology. Combined, SciFig nails both.
7.
Yes. Handwritten labels in any script → clean rendered text.
- Multilingual Handwriting Recognition: Scribble labels in your working language — for example, `récepteur` in French — and SciFig renders them as properly typeset text in the output.
- Bilingual Workflow: Generate the figure in your working language for the lab meeting, one-click swap to English for journal submission — same figure, no rebuild.
- Mixed Labels Supported: English terms, non-English terms, and Greek letters can coexist in a single figure.
8.
Every format your sketch-to-figure output needs for paper, deck, or poster.
- Direct Exports: PNG (recommended for journal submission), JPG (smaller files for sharing or email), and PPTX (a 16:9 PowerPoint slide with the figure embedded — drop straight into a deck).
- Vectorize for Layered Editing: Open the result in Vector Canvas and run Vectorize. One pass produces clean SVG (opens in Illustrator, Inkscape, Figma, or Affinity) plus an editable-shape PPTX where every element is a separately selectable PowerPoint object.
- Resolution: PNG/JPG up to 8K — clears 300 DPI at standard journal column widths.
9.
Yes. Your napkin doodles and whiteboard snaps stay yours.
- Never Training Data: Your hand-drawn diagrams aren't used to train our models — even when they capture brilliant early-stage ideas you haven't published yet.
- Draft-Safe: Whiteboard photos often hold unpublished hypotheses, thesis outlines, or conference-prep sketches. Same encryption-and-zero-training guarantee as any other upload.
- Yours to Delete: Clear your sketch history anytime. Zero retention if you don't want it.