Cycling through all six. Tap any point to stop.

Measured on six things

Turn research ideas into editable scientific figures.

Security

1/12 on the header checklist: no HTTPS enforcement, no HSTS, no CSP, no frame protection, no DMARC. Publication license verification pages are a nice touch but don't substitute for baseline hardening.

Screenshots

1 / 1

Scientific Figure AI

About this app

Scientific Figure AI helps researchers turn text descriptions, reference sketches, and abstracts into scientific illustrations, flowcharts, and graphical abstracts. Users can review a proposed figure plan, then refine labels, arrows, colors, and layouts in the browser. Supported exports include PNG and JPG, SVG with editable vector labels over the illustration, and PowerPoint with editable text and shapes. The site also offers free tools for PRISMA 2020 flow diagrams and phylogenetic trees, plus poster sizing guidance and downloadable research poster templates. AI generation uses credits, with a signup allowance and optional paid plans or credit packs. Researchers should review the scientific content and labels before using any figure in a paper or presentation.

Key features

1AI generation from text, sketches, or abstracts with a reviewable figure plan before drawing.
2Editable vector labels and arrows; exports to PNG/JPG (300/600 dpi), SVG, and editable PPTX.
3Journal palettes (Nature, Science, Cell, Okabe-Ito) and publication license certificates with public verification.
4Free tools: PRISMA 2020 flow diagrams, phylogenetic trees, poster size calculator, and research poster templates.

Target customer profile

Drawn from the product itself, not from a survey.

Demographic

Academic researchers (PhD students, postdocs, PIs)

Pain points

Turning text, sketches, and abstracts into clear scientific visuals takes time.

Primary needs

Editable scientific illustrations, flowcharts, and graphical abstracts for research communication.

Scout Review

Written by AI from measured evidence, scored out of 100.

Scientific Figure AI is a focused, well-positioned tool for academic researchers who need publication-ready figures without a design budget. The homepage does the hard part right — the hero figure sells the output and the inline prompt box lets you try it immediately. Pricing is transparent (100 signup credits, 10 daily, watermark removed on purchase, PRISMA/trees always free), and the differentiators are concrete: editable vector labels, plan-review before generation, journal palettes, PPTX/SVG export, and publication license certificates. The weak spots are execution: keyword-stuffed headings, weak contrast on cards and CTAs, a buried Tools menu, a mobile LCP of 3.8s, and a security posture that's 1/12 on the header checklist. None of these are fatal; all of them are fixable.

Measure by measure

Usability

Clear primary path: inline prompt box and 'Start Creating' both enter the tool, 'Start Free' persists. Pricing explains credits and free tier well. Tools dropdown buries PRISMA/trees/templates.

Design

Competent SaaS template with a genuinely strong hero figure that sells the output. Undermined by keyword-stuffed H2s, weak contrast on pale cards, and washed-out CTA labels.

Speed

Desktop 99/100 is excellent; mobile 87/100 with LCP 3.8s is the drag. CLS 0 and TBT 130ms are clean. Solid but the mobile LCP keeps it out of the top band.

Security

Accessibility

Lighthouse 97/100 with one failing audit (contrast ratio). Language toggle and responsive layout present. No WCAG statement found, so keyboard/screen-reader depth is unverified.

Growth

Real, specific pain (figure creation for papers/posters) with concrete differentiators: editable vector labels, plan-review, journal palettes, PPTX/SVG, license certificates. Early stage (domain 0.0 yrs).

Improvement areas

Security headers: enforce HTTPS and add HSTS, CSP, frame protection, X-Content-Type-Options, Referrer-Policy, and DMARC.
Mobile performance: LCP 3.8s drags mobile Lighthouse to 87; compress hero assets and cut render-blocking resources.
Copy and contrast: rewrite keyword-stuffed H2s and fix weak contrast on pale cards and the cyan gradient CTA.
Navigation: surface PRISMA, trees, and poster templates at top level instead of burying them in a Tools dropdown.

Design & Usability

The homepage does the hardest thing right: the hero figure (a labelled CAR-T cell killing a tumor cell) immediately shows what the product produces, and the inline prompt box turns the landing page into a working entry point. That's better than most SaaS brochures. The problems are execution details: nearly every H2 repeats 'AI Scientific Figure Generator' like a keyword bingo card, small light-grey body text on pale gradient cards is hard to read, and the white label on the cyan end of the primary CTA washes out. Navigation is minimal and scannable, but the Tools dropdown buries PRISMA, phylogenetic trees, and poster templates behind one click — those are free-tool lead magnets and they deserve top-level visibility. Pricing is unusually clear: 100 signup credits, 10 daily check-in, watermark removed on any purchase, and PRISMA/trees always free.

Speed & Security

Speed is a tale of two devices. Desktop Lighthouse hits 99/100, but mobile lands at 87 with an LCP of 3.8s — above the 2.5s 'good' threshold and the main thing keeping this out of the top band. CLS 0 and TBT 130ms are clean, so the fix is likely image weight or render-blocking on the hero figure. Security is the weak spot: the HTTP probe found HTTPS not enforced, no HSTS, no CSP, no frame protection, no X-Content-Type-Options, no Referrer-Policy, no Permissions-Policy, and no DMARC — 1/12 on the checklist. For a tool handling research IP and issuing publication license certificates with public verification pages, that's a gap worth closing before pitching labs. No SOC 2, ISO, pen-test, or bug-bounty signals were found in the collected evidence.

Accessibility & Growth

Accessibility scores 97/100 on Lighthouse with a single failing audit: insufficient contrast between background and foreground colors — the same pale-card issue the visual audit flagged. A language toggle and responsive layout are present, but no WCAG statement was found, so keyboard and screen-reader depth is unverified. On growth: the target customer is specific (PhD students, postdocs, PIs making figures for papers and posters), the pain is real and recurring, and the differentiators are concrete — editable vector labels instead of flat images, a plan-review step before generation, journal palettes (Nature/Science/Cell/Okabe-Ito), PPTX and SVG export, and publication license certificates. The category has a clear incumbent (BioRender, which the site itself name-checks) and room for a cheaper AI-native entry. The domain was registered 2026-11-10, so this is very early — traction is context, not the score. The wedge is real; the question is whether it can out-execute an entrenched incumbent on trust and library depth.

Conclusion

If you're a researcher who's spent an afternoon dragging icons in a vector editor and still ended up with a figure a reviewer will squint at, this is worth a free account. The plan-review step and editable labels are the right ideas, and the free PRISMA and phylogenetic tools are a smart on-ramp. Before you trust it with a manuscript figure, though: check the labels yourself (the site says so too), and don't expect enterprise-grade security signals — there aren't any yet. For an early-stage tool, the bones are good. The polish and the hardening are what's missing.

Suggestions for improvement

Enforce HTTPS and add HSTS, CSP, X-Content-Type-Options, Referrer-Policy, and DMARC — 1/12 on the header checklist is a trust problem for a research tool.
Fix mobile LCP (3.8s): compress or lazy-load the hero figure and audit render-blocking resources to push mobile Lighthouse above 90.
Rewrite H2s to describe actual sections instead of repeating 'AI Scientific Figure Generator'; fix contrast on pale cards and the cyan gradient CTA.
Surface PRISMA, phylogenetic trees, and poster templates at top level — they're free lead magnets currently buried in a Tools dropdown.

How we score

Alternatives comparison

Named competitors, point by point. Nobody paid to appear here or to be left out.

	Scientific Figure AI
AI figure generation from text/sketch/abstract	Yes — text, sketch, or abstract with plan review	Not specified in provided details
Editable vector labels and arrows	Yes — labels rebuilt as real text, arrows as vector objects	Not specified in provided details
Export formats	PNG/JPG (300/600 dpi), SVG, editable PPTX	Not specified in provided details
Free tier	100 signup credits + 10 daily check-in; watermarked exports	Not specified in provided details
Paid pricing	Starter $84/yr, Pro $180/yr, Lab $468/yr (yearly)	Not specified in provided details
Free PRISMA / phylogenetic tools	Yes — always free, no credits	Not specified in provided details
Publication license certificate	Yes — numbered certificate with public verification page	Not specified in provided details