Start with any input
- YOU
- Drop text, a sketch, a reference figure, a PDF page, or a lab photo.
- SCIFIG
- Generates a journal-grade draft in one of 6 publication styles — or let Auto decide.
- YOU GET
- A publication-ready first draft in seconds.
AI Scientific Illustrator
Turn text, sketches, references, PDFs, and photos into Scientific Figures
Editable at every step — describe your scientific figure in plain language, choose from multiple generation modes, and fine-tune every label and layout along the way.
Core Subject (e.g., Cas9 protein cutting DNA)
Action / Details (e.g., Double strand break, detailed molecular view)
Get 200 free credits | No credit card required
6 input modes · 6 publication stylesEvery text editable · Multimodal enhanceEditable PPTX · Layered SVG · 8K PNG / JPG
Scientific Figures Made with SciFig
Explore publication-ready figures created by researchers worldwide.
SciFig — one tool, three stages, zero rework.
Generate, enhance, and vectorize your scientific figures in one place.
GENERATE
Turn any input into a publication figure
From text, sketch, reference, PDF, or photo, SciFig generates your figure in one of six publication styles — scientific, line, 3D, editorial, sketch, or watercolor — engineered to match Nature, Cell, and Science. Auto mode skips the style overlay and follows your prompt alone.
- Six input modes
- Six publication styles + Auto mode
- Journal-grade output
ENHANCE
Refine every detail without starting over
Unlike one-shot AI, SciFig keeps everything editable after generation — change any text's content, font, size, or position directly, select a region and regenerate it with a natural-language prompt, or feed a PDF as context for smarter edits.
- Every text editable
- Precision region inpaint
- Multimodal + 8K upscaling
VECTORIZE
Take full control of every detail
Export figures as editable PPTX or layered SVG, then fine-tune every shape, color, and label in SciFig's built-in canvas with full undo/redo. No Illustrator or Figma roundtrip needed.
- Editable PPTX
- Layered SVG + undo/redo
- Built-in canvas
From capability to the SciFig canvas
Generate, Enhance, and Vectorize — here's what they each render to.
How it works
Three steps inside SciFig. Editable at every one.
Refine without starting over
- YOU
- Click any label to retype it. Circle any region to regenerate it. Attach a PDF for context.
- SCIFIG
- Edits text in place, inpaints regions, and upscales to 8K — no full re-render needed.
- YOU GET
- A scientific figure that matches your data, not a lookalike.
Deliver, still editable
- YOU
- Export to editable PPTX, layered SVG, or 8K PNG / JPG.
- SCIFIG
- Keeps every shape, label, and color editable in the built-in canvas.
- YOU GET
- Journal-, poster-, and talk-ready scientific figures — no Illustrator roundtrip.
Trusted by researchers
What scientists actually say
10x
Faster scientific figure creation
5
AI generation modes
4+
Export formats
“I used to spend an entire day redrawing scientific figures for revision. With SciFig I describe the change in one sentence and get a publication-ready scientific figure in minutes.”
Dr. Wei L.
Postdoc, Molecular Biology, Stanford
“The editable PPT export is a game-changer. I can fine-tune labels and colors right in PowerPoint instead of going back to Illustrator every time a reviewer asks for changes.”
Prof. James R.
Associate Professor, Biomedical Engineering, MIT
“Our lab switched from BioRender last semester. The layered SVG output and built-in canvas give us far more control, and the AI actually understands scientific context.”
Maria S., PhD
PhD Candidate, Neuroscience, UCL
“I run a lab of 15 people. Before SciFig every student came to me for scientific figure help. Now they generate journal-quality scientific figures on their own — I just review and approve.”
Prof. Yuki T.
Principal Investigator, Systems Biology, ETH Zürich
“As a journal editor I see hundreds of scientific figures a month. Papers with SciFig scientific figures consistently have cleaner layouts, correct labeling, and proper resolution. It shows.”
Dr. Elena V.
Associate Editor, Cell Reports Methods
“The PDF-to-Figure mode saved our grant deadline. I uploaded our 40-page proposal and got presentation-ready scientific figures for every key concept in under an hour.”
Kevin C., PhD
PhD Candidate, Computational Chemistry, Caltech
Trusted by 100K+ researchers from leading institutions:
FAQ
Everything you need to know about SciFig
Common questions about creating scientific figures with SciFig.
SciFig is a free online AI scientific figure illustrator. From any input to a journal-ready scientific figure — in minutes, not days.
- Built for Science Work: Researchers, students at any level, educators, industry R&D teams, and science communicators — anyone turning complex science into a clear figure, without wrestling with Illustrator.
- Any Input Welcome: Type a description, drop a PDF, snap a lab photo, or upload a reference image. SciFig turns all four into the same clean, publication-ready figure.
- Figures Journals Accept: Mechanism diagrams, signaling pathways, experimental workflows, medical illustrations, molecular structures — all print-ready for peer-reviewed journal submissions, whatever your target publication.
BioRender gives templates. Raw AI gives flat pixels. We give science-grade figures you can fully edit.
- End-to-End, All-in-One: BioRender is a fixed template library — when your topic isn't in it, you're stuck. SciFig generates one-of-a-kind figures from your text, PDF, photo, or sketch, then exports as SVG or PPTX so you can keep polishing in PowerPoint, Illustrator, or Inkscape.
- Editable Vectors, Not Flat AI Pixels: Nano banana and Midjourney bake a mislabeled protein into pixels you can't edit — one wrong enzyme is a career-ending reviewer comment. SciFig outputs every element as a separate SVG layer. Click, type, fixed.
- Free Trial, $12 Entry: Try every feature on free credits first — no credit card needed. When you upgrade, paid plans start at just $12/month (annual) or $18/month (monthly) — whichever fits your research cadence.
Yes. SciFig isn't a bio-only tool — 11 disciplines supported natively, with published examples you can browse and verify.
- Far Beyond Biology: Biology, chemistry, physics, engineering, computer science, energy, ecology, bioengineering, astronomy, agriculture, materials — the major disciplines are all in. Mechanisms, workflows, cross-sections, ecosystems, graphical abstracts, journal covers, lab apparatus, microstructures, systems & networks — all nine high-frequency figure types covered.
- No More "Bio-Only" Tools: BioRender locks its asset library to molecules and cells — try drawing a semiconductor band diagram, satellite orbit, or catalytic cycle and you're stuck if it's not in the template. SciFig generates straight from your description, PDF, or sketch — no fixed library required.
- See for Yourself: Open Inspiration, filter by discipline, and every example ships with its original prompt — copy it into SciFig and reproduce it yourself.
That's exactly who we built this for. SciFig is the AI scientific illustrator for the blank-canvas problem.
- Describe It in Plain Language: Use Text-to-Figure. Type "CRISPR cutting DNA with PAM site" and SciFig proposes multiple layouts based on how peer-reviewed journals have visualized similar concepts.
- Let a Paper Brief the AI: Use PDF-to-Figure. Upload the paper you're working from — SciFig's LLM reads it and generates a new figure informed by the concepts inside. Stop staring at a blank canvas.
- Start from Real Life: Use Photo-to-Figure or Sketch-to-Figure. Snap your lab rig or a whiteboard doodle — SciFig turns it into clean vectors. No design experience, no template library needed.
AI gets you 99% there. You handle the final 1% — and that's the whole point.
- Every Element is Editable: Labels, arrows, molecules, organs, legends — each a separate object. Click to fix a mislabeled protein, swap a wrong enzyme, or re-route an arrow — one click, done.
- 99% = 0% in Science: Raster AI tools (Midjourney, nano banana) bake mistakes into pixels you can't edit. SciFig keeps every element editable because you — the expert — are the final authority.
Yes. You own every figure — for journals, for products, for anything legal.
- Full Publication Rights: Paid plans come with full authority to publish in peer-reviewed journals, on preprint servers, and in your thesis. No tier-gating on publication.
- Full Commercial Rights: Use your figures in textbooks, commercial products, courses, consulting decks — no additional licensing fees, no revenue share, no "just for research" fine print.
- Zero Ownership Claim: SciFig claims no ownership, no usage license, no right to feature your work. (Note: Free plans are for non-commercial evaluation only — upgrade before publication or commercial use.)
It's your call — and your target journal's policy. No universal rule yet.
- Check Your Target Journal First: AI-disclosure policies change fast and vary by publisher. Spend 30 seconds on your journal's author guidelines before submission — it's almost always answered there.
- When in Doubt, a Methods Line Covers You: If the policy is ambiguous, a one-line "Figures prepared with SciFig (AI-assisted)" in the methods section is the safe middle ground — no over-explaining, no hiding.
Every format journals, posters, and slides actually need.
- Editable Vector: SVG — open and keep editing in Illustrator, Inkscape, Figma, or Affinity Designer.
- Editable Slides: Native PPTX — every shape is a separate PowerPoint object. Tweak labels, resize panels, recolor without round-tripping to another tool.
- Journal-Ready Raster: PNG and JPG up to 8K. Meets standard journal submission guidelines out of the box.
Yes, absolutely. 1K and 2K hit 300 DPI at standard journal column widths.
- Single-Column Figures: 1K resolution hits 300+ DPI at standard single-column width — cleared for submission out of the box.
- Double-Column Figures: 2K resolution hits 300+ DPI at full-page width. No reviewer "please improve figure resolution" comments.
- Posters & Covers: Need bigger? One-click Upscale to 4K/8K for A0 posters and journal cover art — detail stays sharp, no pixelation.
Seconds per draft. Minutes end-to-end.
- Initial Draft: Seconds from prompt to first draft — faster than filling in a BioRender template.
- Multi-Round Refinement: Each iteration lands in seconds. Lock what works, regenerate just the panel that doesn't.
- End-to-End: A few minutes from blank canvas to export-ready figure — including your manual edits and the final format pick.
Yes. Invoice-ready, grant-line-item-ready, zero finance friction.
- Formal Invoices: We issue tax-compliant invoices with your institution's address, tax ID, and itemized usage — drop them straight into your reimbursement system.
- Grant Line Item: SciFig bills cleanly as "scientific visualization software" — the same category your lab probably already covers for GraphPad, Prism, or Illustrator.
- Two Subscription Shapes: Monthly and annual subscriptions (for regular figure makers) and team / institution plans with bulk credits and single-invoice billing. Contact us for a lab or department quote.
Yes. Your unpublished research stays yours, full stop.
- Zero AI Training: We NEVER use your private uploads, prompts, or generated figures to train our models — ever.
- Encrypted End-to-End: All files are encrypted in transit and at rest. Nobody else can see your work.
Ready to publish
Scientific figures. Edit anything. Publish today.
Start for FreeFree to start · No credit card required · Trusted by researchers worldwide
Text-to-FigureSketch-to-FigureReference-to-FigurePDF-to-FigurePhoto-to-Figure6 Publication StylesText-to-FigureSketch-to-FigureReference-to-FigurePDF-to-FigurePhoto-to-Figure6 Publication StylesText-to-FigureSketch-to-FigureReference-to-FigurePDF-to-FigurePhoto-to-Figure6 Publication Styles
Every Text EditablePrecision InpaintMultimodal Enhance8K UpscalingEditable PPTXLayered SVG8K PNG / JPGEvery Text EditablePrecision InpaintMultimodal Enhance8K UpscalingEditable PPTXLayered SVG8K PNG / JPGEvery Text EditablePrecision InpaintMultimodal Enhance8K UpscalingEditable PPTXLayered SVG8K PNG / JPG