Face Gallery API
The face gallery manages enrolled faces for you — handling storage, indexing, and similarity search — so you can recognize people across new images without building your own vector database or matching pipeline.
Gallery Actions
- Enroll — Register a face with an external ID (your user ID, employee number, etc.). The gallery stores the face embedding and links it to your identifier.
- Search — Submit a new face image and receive ranked matches from the gallery, ordered by similarity score.
- Compare — 1:1 check between a new face and a specific enrolled identity. Returns a similarity score and a match/no-match decision at your chosen threshold.
- List — Retrieve all enrolled faces with their external IDs and enrollment timestamps.
- Delete — Remove an enrolled face from the gallery by its external ID.
When to Use the Gallery vs. Embeddings
Use the gallery when you want Mental Edge to handle persistence and similarity search. Use the raw embeddings endpoint when you need to store vectors in your own infrastructure (Postgres, Pinecone, pgvector, etc.) or apply custom matching logic. Both approaches use the same underlying face model.
Priced at the Face Recognition rate ($0.001 per gallery search or compare) in production. Free sandbox available.
Face Embeddings (bring your own storage) | API Pricing | Developer API overview