Integrate ScanSkinAI
Deploy clinical-grade skin analysis in minutes — 2-line embeddable widget, REST API or SSO. Built by clinicians, with 95.3% concordance and enterprise-ready security.
Insurer or employer instead? See ScanSkinAI Business Solutions →
Two very different integrations. Pick what fits.
The Embeddable Widget is a ready-made UI you drop onto your site in 2 minutes — no design, no backend. The REST API is for teams building their own custom product, UI/UX, and workflows around the AI.
Embeddable Widget
The fastest path to production. A pre-built, high-performance UI component that handles image capture, analysis and reporting — no backend required.
- Live in 2 minutes — just 2 lines of HTML
- No backend, no build tools, no design work
- UI, dark mode, EN/ZH/ES — all built-in
- Best for CMS, marketing sites, pharmacies
REST API
Maximum flexibility. Direct access to our clinical inference engine via secured JSON endpoints, webhooks and SSO — you own the UI and workflow.
- Days to weeks — you design the UI/UX
- Multi-lesion detection, triage, webhooks
- SSO / OAuth2 + full data-residency control
- Best for telehealth, clinic platforms, insurers
Embeddable Widgets — 2 lines of code
Add AI-powered skin analysis to any website in under 2 minutes. Choose the general skin-check widget (80+ conditions, confidence scoring and plain-English summaries) or the facial-scan widget (12 facial metrics + personalised routine recommendations). Both are plug-and-play Web Components — no backend, no build tools.
<!-- Add once to your page --> <script src="https://www.scanskinai.com/widget/v2/scanskinai-widget.js"></script> <!-- Place anywhere on your site --> <scanskinai-widget api-key="YOUR_SECRET_KEY"></scanskinai-widget>
No npm. No backend. No build process. Works on any website or CMS.
Facial scan widget also available
For beauty brands, salons and skincare stores, use the facial-scan widget to let customers analyse 12 facial metrics and get a routine recommendation in 90 seconds. Same two-line embed, no backend required.
<!-- Add once to your page --> <script src="https://www.scanskinai.com/widget/facial/v1/scanskinai-facial-widget.js"></script> <!-- Place anywhere on your site --> <scanskinai-facial-widget api-key="YOUR_SECRET_KEY"></scanskinai-facial-widget>
See it live — this is what those 2 lines render
Click any state to preview the full user journey, from upload to AI result.
For informational purposes only. Not a medical diagnosis. Always consult a qualified dermatologist.
Upload / idle — awaiting photo
Zero dependencies
Pure Web Component, isolated from your site's CSS and JS — no conflicts.
Works everywhere
WordPress, Webflow, Shopify, Wix, React, or plain HTML — all supported.
Choice-based inputs
Body location, duration, symptoms and progression — pre-built dropdowns, no typing required.
Dark mode built-in
Automatically follows your visitor's device preference.
Multi-language
English, Chinese (ZH) and Spanish (ES) out of the box.
Fully accessible
ARIA roles, keyboard navigation and screen reader support.
Brand-flexible
Custom accent colour and optional white-label mode.
Domain-locked keys
Your API key only works on your registered website — secure by design.
Privacy-first
ScanSkinAI does not store uploaded images or scan results.
Getting started in 3 steps
Activate your domain
Email info@scanskinai.com with your deployment domain so we can activate the key.
Paste & publish
Paste the 2 lines of code into your website and publish. Done.
Build your own product on top of the ScanSkinAI API
For teams that need full control of the UI/UX, data flow and clinical workflow. Returns structured lesion classification, probability, multi-lesion detection and clinical triage. Requires backend and frontend development.
Built for the organisations powering modern skin health
Six customer segments, one unified ScanSkinAI platform.
Health & Beauty Apps
Consumer skincare, wellness and beauty apps embedding AI skin analysis into their own experience.
Telemedicine Platforms
Virtual care and online dermatology services automating triage and pre-consult assessments.
Skin Product Websites
DTC skincare brands and marketplaces personalising recommendations with data-driven matching.
Healthcare Providers & Workers
Clinics, GPs, dermatology specialists and researchers integrating AI into EHR / EMR workflows.
Laboratory Service Providers
Diagnostic labs and research groups using AI skin data to enrich testing and reporting pipelines.
Online Pharmacies
Pharmacies and pharma retailers closing the loop from skin scan to prescription and product sale.
One REST API, three industries served
One ScanSkinAI skin lesion analysis API — used across consumer apps, clinics and product ecosystems.
Connect with your customers
Call the ScanSkinAI REST API from your app to deliver clinically informed skin lesion analysis, personalised follow-ups and health guides — all under your brand.
Send an image, receive structured JSON in seconds. The API returns lesion classification, confidence and a risk score straight back to your app — every interaction stays inside your own experience.
Telemedicine ready
Automate your online dermatology consultations with AI-powered triage and structured pre-consult reports your clinicians can sign off in seconds.
REST API integration
One HTTPS endpoint — works with any backend or mobile stack.
Personalised follow-ups
Use structured results to drive re-scan reminders and content.
Health & education content
Deliver tailored skin guides that drive engagement and retention.
Telemedicine automation
Pre-screen patients before video consults to save clinician time.
How integration works
Integration is REST API only — no SDKs to install, no EHR connectors required. Works with any backend: Node, Python, PHP, Ruby, Go or .NET.
Apply for Sandbox
Sign up at ivyai.hk to instantly receive sandbox API credentials — no sales call required.
Build & Test
Call our REST endpoints with your sandbox key. Full request / response samples, no rate caps for development.
Go Live
Submit a short production review (~24h). Receive your production key and start serving real users.
One endpoint. Comprehensive lesion analysis.
Every call returns clinic-validated structured data — classification, confidence, probability and triage in one response.
Lesion Classification
Identify moles, skin tags, rashes and suspicious lesions with a primary class label.
Confidence Scores
Model confidence for the predicted class so you can set your own review thresholds.
Probability (0–100)
Probability the lesion matches the predicted condition — drive in-app prioritisation and follow-ups.
Clinical Triage
Four-level triage recommendation: self-monitor, GP visit, dermatologist, urgent.
Differential Diagnosis
Top alternative conditions returned with each call to support clinical review.
Multi-language Output
Recommendations available in EN, ZH, JA, KO, ES, FR, DE and more.
Multi-Lesion Detection
Detect, locate and classify every lesion in a single full-body or regional photo — one API call returns all bounding boxes.
Your first scan in one HTTP request
Any HTTP client works — no SDK required. Below are copy-paste examples in cURL, Node.js and Python, plus a real sample JSON response.
cURL
curl https://api.scanskinai.com/v1/scans \ -H "Authorization: Bearer $SCANSKINAI_KEY" \ -H "Content-Type: multipart/form-data" \ -F "image=@mole.jpg" \ -F "body_location=back" \ -F "language=en"
Node.js (fetch)
const form = new FormData();
form.append("image", fs.createReadStream("mole.jpg"));
form.append("body_location", "back");
const res = await fetch("https://api.scanskinai.com/v1/scans", {
method: "POST",
headers: { Authorization: `Bearer ${process.env.SCANSKINAI_KEY}` },
body: form,
});
const data = await res.json();Python (requests)
import os, requests
r = requests.post(
"https://api.scanskinai.com/v1/scans",
headers={"Authorization": f"Bearer {os.environ['SCANSKINAI_KEY']}"},
files={"image": open("mole.jpg", "rb")},
data={"body_location": "back", "language": "en"},
)
data = r.json()Sample response (HTTP 200)
{
"request_id": "scn_01HK9V…",
"model_version": "skin-v4.2",
"primary": {
"class": "melanocytic_nevus",
"label": "Benign mole",
"confidence": 0.94,
"probability": 0.91
},
"differential": [
{ "class": "seborrheic_keratosis", "confidence": 0.04 },
{ "class": "melanoma", "confidence": 0.02 }
],
"triage": {
"level": "self_monitor",
"recommendation": "Re-scan in 3 months. See a GP sooner if the mole changes.",
"language": "en"
},
"quality": { "focus": "ok", "lighting": "ok", "is_skin": true },
"latency_ms": 1832
}Full reference — endpoints, error codes, webhooks and Postman collection — ships with your sandbox credentials.
Fast, predictable monthly plans
Transparent rate limits, sub-2-second median latency and monthly call plans that scale with your volume.
< 2s median latency
Single-lesion classification from EU / UK regions. 4–8s for multi-lesion full-body scans.
99.9% uptime SLA
Enterprise contracts include credit-back terms and prioritised routing.
120 req/min production
Sandbox 60 rpm. Burst up to 300 rpm for 10s. Enterprise keys get custom limits.
Plans from $99 / month
Free 50-call sandbox, then Starter / Growth / Scale monthly plans. Enterprise custom.
Multi-Lesion Detection API
Detect, locate and classify every lesion in a single full-body or regional photo — one API call returns per-lesion bounding boxes, predicted condition, confidence and quality flags.
What every response includes
- Per-lesion bounding box coordinates (x, y, w, h)
- Predicted class: melanoma, nevus, BCC, seborrheic keratosis, normal skin, and more
- Confidence score per lesion (0–100%)
- Stable lesion ID for longitudinal tracking across follow-up scans
- Quality flags: "too small", "not a skin image", out-of-focus
Ideal for
Teledermatology triage · Full-body mole mapping platforms · Clinical research cohorts · Insurer and employer screening programmes.
Interested in early access?
Currently in private beta — pricing and rate limits provided on request.
Security & compliance
Built for regulated industries. Audit-ready out of the box.
Frequently asked questions
Pricing, latency, rate limits, data retention and compliance — the answers engineering and procurement teams ask most.
The sandbox is free forever for development and testing. Production is priced per successful scan on a tiered volume model: Starter (pay-as-you-go from $0.20/scan), Growth (bulk-committed monthly volume with lower per-scan pricing) and Enterprise (custom volume, SLA and data-residency terms). The embeddable widget is billed the same way as the REST API — one successful classification = one scan. Email info@scanskinai.com for a written quote with your monthly volume.
Median end-to-end latency for a single-lesion classification call is under 2 seconds from EU and UK regions, and typically 3–4 seconds globally. Multi-lesion full-body detection returns in 4–8 seconds depending on image resolution and lesion count. Enterprise contracts include a 99.9% monthly uptime SLA with credit-back terms, prioritised routing and a status page subscription.
The API accepts JPEG, PNG, HEIC and WebP up to 20 MB per image. Recommended minimum is 1024×1024 for single-lesion scans and 2048×1536 for multi-lesion full-body scans. Images are auto-oriented from EXIF, resized server-side and quality-flagged — if a photo is out of focus, too dark, or not a skin image, the API returns a structured quality error instead of a low-confidence prediction.
By default, the REST API does not persist uploaded images — they are processed in memory and discarded after the response. Scan metadata (timestamp, class, confidence, request ID) is retained for 30 days for debugging and rate-limit enforcement, then purged. Enterprise customers can opt in to longer retention with EU/UK data-residency, or fully disable metadata retention under a signed DPA.
ScanSkinAI operates under ISO 27001:2022 information-security management and UKCA Class I medical-device registration. The service is GDPR-aligned by default; HIPAA-ready deployments are available with a signed BAA for US customers. All traffic is TLS 1.2+ and API keys are bearer tokens scoped to a single environment (sandbox or production) and, for widgets, to a single registered domain.
Sandbox: 60 requests/minute per key, no daily cap, intended for development. Production: 120 requests/minute per key by default with burst allowance up to 300/minute for 10 seconds. Enterprise keys are provisioned with custom per-second and per-minute limits — including dedicated capacity for full-body multi-lesion workloads.
The REST endpoints are language-agnostic — any HTTP client works. We publish tested code samples in cURL, Node.js (fetch/axios), Python (requests), PHP, Ruby, Go and .NET. No official SDK is required; a lightweight TypeScript helper is available on request. Response messages (triage recommendations, plain-English summaries) are available in EN, ZH, JA, KO, ES, FR and DE via the Accept-Language header.
Server-to-server is the recommended pattern — your backend proxies the request so the API key stays private. For browser-only use cases, use the embeddable widget with a domain-locked publishable key. For native iOS or Android apps, embed the call in your own backend or use a short-lived signed URL provisioned by your server; never ship a production API key inside a mobile binary.
Headquartered in the heart of Canary Wharf
ScanSkinAI is built by Ivy AI Solutions Limited, a London-based health-tech startup headquartered at One Canada Square in Canary Wharf. We design our workplace skin cancer screening platform for UK small businesses, with UKCA Class I registration and ISO 27001 information security.
From our London HQ we support partners across the United Kingdom, Hong Kong, Singapore and the Asia-Pacific region — with clinical validation, enterprise security and dedicated integration support.
Start building today
Apply at ivyai.hk and receive your sandbox API key instantly. Production access is reviewed within ~24 hours.
Sandbox is free. Production access after review (~24h).
See API pricing · Not a developer? See Brand solutions or Spa solutions.