CompreFace is a free, open-source face recognition system from Exadel that packages detection, verification, and identification into a self-hosted service with a REST API. It has earned more than 8,300 stars on GitHub and is built on proven models such as FaceNet and InsightFace. The project was last updated in late 2024 and remains a popular self-hosted alternative to cloud face APIs.
CompreFace is a face recognition service that can be integrated into any system without prior machine learning skills. It provides a REST API for face recognition, face verification, face detection, landmark detection, mask detection, head pose estimation, and age and gender recognition, and it deploys in minutes with Docker. An administration UI lets you manage face collections, users, and API keys without writing code, while the underlying recognition plugins can be swapped depending on the accuracy and speed trade-offs you need. Because everything runs on your own infrastructure, face images never leave your network, which is a deciding factor for organizations that cannot send biometrics to third-party clouds. The codebase uses Docker Compose for a one-command start.
CompreFace is free and open source under the Apache-2.0 license. There is no per-call pricing and no cloud subscription: you download it, run it with Docker, and pay only for your own hosting. Exadel offers separate commercial services around its accelerators, but the software itself costs nothing.
CompreFace is for developers and organizations that need face recognition under their own control: access-control and attendance systems, photo management apps, and identity verification flows where data-residency rules rule out cloud APIs. It is especially attractive when you want API-style convenience without the per-request bill.