Send an ingredient list. Get an efficacy grade with the receipts.

Quickstart

Grade any ingredient list — no key needed to evaluate:

# GET — paste an INCI string (percentages optional but they drive the grade)
curl "https://visagetheory.com/api/grade?inci=Niacinamide 10%25, Zinc PCA 1%25, Glycerin"

# POST — structured actives
curl -X POST "https://visagetheory.com/api/grade" \
  -H "Content-Type: application/json" \
  -d '{"actives":[{"name":"Vitamin C","percent":15},{"name":"Vitamin E","percent":1},{"name":"Ferulic Acid","percent":0.5}]}'

Response:

{
  "letter": "A",
  "score": 93,
  "summary": "Hero actives are dosed to clinical levels...",
  "evidenceLevel": "formulation",
  "recognizedActives": [ { "name": "Vitamin C", "percent": 15, "status": "strong" }, ... ],
  "receipts": [ { "headline": "...", "detail": "..." } ],
  "disclaimer": "...predicts likely efficacy — it does not test the product...",
  "attribution": "Efficacy grade by Visage Theory."
}

What you can build

• 🏷️ PDP badges — show an independent efficacy grade on your product pages (with a link to the receipts).
• 🔬 R&D pre-checks — grade a formula before you commit to a production run; see which active is underdosed.
• 🛒 Retail filters — let shoppers sort a category by what actually works, not by margin.
• 🤖 Agents & tools — give an assistant a real efficacy answer with a citable source.

Integrity is enforced server-side: an elevated evidence tier is only honored with a resolvable study/COA URL, so the API can't be used to manufacture an unearned grade. See the open methodology.