Methodology last reviewed: July 2026
AuraMed reads the label you scan (or the name you type or say), then looks for supporting records in public reference sources before explaining the result in plain language. Results are generated with the help of an AI provider (Anthropic) and are educational — they are not independently verified medical judgement.
Medicines: the U.S. FDA public drug-label API (openFDA). AuraMed does not currently query any non-U.S. medicines database.
Supplements: the NIH Dietary Supplement Label Database (DSLD).
Cosmetics: no official cosmetic-label database is connected. A cosmetic result describes only the ingredient text transcribed from the photo you scanned, and is labelled as unmatched.
Name search: the NIH RxTerms API powers the medication-name suggestions.
When a result is matched to an official record, AuraMed shows a source badge naming that record. When no reliable record is available, the result is clearly labelled as unmatched. AuraMed then describes only the label text it read and asks you to verify the physical package — it does not fill the gap from general knowledge. Coverage varies — not every product sold has an official label record.
Each scan carries a confidence indication. Low-confidence identifications are not analyzed — AuraMed asks you to rescan or check the label instead of guessing. Interaction and timing output is presented as potential considerations to review, never as definitive instructions.
AuraMed is built and reviewed by a licensed pharmacist. Further credential details will be published here once approved for publication.
Every result includes a report action, or email auramed.support@gmail.com with the product name and a photo of the label.