CompreFace
github.com
- Free plan
- Open source
- Windows
- macOS
- Linux
- Web
- API
- Self-hosted
What CompreFace is
CompreFace is a free, open-source face recognition service that provides its functions through a REST API. It supports face recognition and verification, plus face, landmark, and mask detection, head pose detection, age recognition, and gender recognition. The project says users can set it up without machine-learning experience. It is delivered as a Docker Compose configuration and can run as a standalone server or be deployed in the cloud; the listed deployment options also include hybrid use. Setup documentation covers Windows, macOS, and Linux and requires Docker and Docker Compose or Docker Desktop. Hardware requirements include an x86 processor with AVX support, and CPU and GPU models are supported. A UI panel and role-management system let users manage roles and access to recognition services. The project names FaceNet and InsightFace among its libraries and links JavaScript, Python, and .NET SDK repositories. CompreFace is self-hosted and released under Apache 2.0. Its quick start uses a release archive and the command docker-compose up -d to start the service.
Who it suits
It suits developers or teams that need face-related capabilities through an API and can manage a Docker-based deployment. The project says prior machine-learning skills are not required.
Good to know, and what to watch
- Free and open-source under Apache 2.0.
- Supports recognition, verification, and multiple detection tasks.
- CPU and GPU models are supported.
- Role management controls access to recognition services.
- Setup requires Docker and Docker Compose or Docker Desktop.
- Hardware requirements include an x86 processor with AVX support.
- Quick start involves downloading an archive and running a command.
Plans
Every plan and priceCompreFace
Free
Free plan
Getting started
- Open the CompreFace GitHub website
- Use Windows, macOS, or Linux with Docker and Docker Compose or Docker Desktop
- Download the release archive
- Start the service with docker-compose up -d
- Connect through the REST API or linked SDKs
Limits to check
The free CompreFace plan is Docker-based and requires suitable CPU or GPU hardware, including an x86 processor with AVX support. Deployment is self-hosted rather than a managed service.
The verdict
CompreFace offers a broad face-analysis API in a self-hosted package, with CPU and GPU model support. Its Docker setup and stated processor requirement are important to check before deployment.
Questions people ask
Is CompreFace free?
Yes. The listed CompreFace plan costs nothing and is free and open source.
Which platforms are supported?
The service supports API, Linux, macOS, self-hosted, web, and Windows use. Setup instructions cover Windows, macOS, and Linux.
Does it offer an API?
Yes. CompreFace provides a REST API for its face-recognition and detection capabilities.
Is CompreFace open source?
Yes. It is released under the Apache 2.0 license.
What SDKs are available?
The project links SDK repositories for JavaScript, Python, and .NET.
What the maker publishes
- Access management
- A UI panel and role-management system let users manage roles and control access to face recognition services. source
- Access management
- The web interface includes roles and permissions for controlling access to face services. source
- Age recognition
- An age-recognition plugin is listed among the available services. source
- API documentation depth
- The repository documentation includes endpoint descriptions for subjects, examples, recognition, verification, detection, embeddings and base64 requests. source
- Cloud and on-premises
- Official feature documentation says the software can be deployed in the cloud or on premises. source
- Company headquarters and founding
- LinkedIn lists Walnut Creek, California as headquarters and 1998 as the founding year. source
- CPU and GPU
- The project documents model variants that run on both CPU and GPU hardware. source
- Data-control positioning
- The project presents self-hosting and on-premises deployment as ways to keep face data under the operator’s control. source
- Deployment
- It is delivered as a Docker Compose configuration and can run as a standalone server or be deployed in the cloud. source
- Deployment model
- CompreFace is distributed as Docker Compose configuration and can run as a standalone server or in cloud environments. source
- Documentation scope
- The documentation index also covers configuration, architecture and scalability, custom builds, migration, roles, plugins, Kubernetes and installation options. source
- Extensibility
- The project invites contributors to add SDKs, plugins and integrations such as Home Assistant or DreamFactory. source
- Face detection
- A face-detection service is included alongside recognition and verification. source
- Face recognition
- The service can identify enrolled people from submitted images. source
- Face verification
- The API supports comparing faces to verify whether they belong to the same person. source
- Gender recognition
- A gender-recognition plugin is available. source
- Hardware
- CompreFace supports CPU and GPU models and lists an x86 processor with AVX support among its requirements. source
- Independent deployment example
- A 2026 technical article used CompreFace with Immich, Home Assistant, UniFi Protect and Double Take in a self-hosted doorbell workflow. source
- Integrations
- The project invites contributors to add support for other platforms such as Home Assistant or DreamFactory; it does not describe these as built-in integrations on the page. source
- Landmark detection
- A plugin provides facial-landmark detection. source
- Legal name and owner
- Crunchbase identifies the legal entity as Exadel, Inc. and says it was acquired by Sun Capital Partners. source
- Licence
- CompreFace is released under the Apache 2.0 open-source licence. source
- Limit: Docker operational burden
- An independent project report said CompreFace required multiple Docker instances, complicating deployment on machines without Docker or docker-compose. source
- Limit: ecosystem maintenance
- A Home Assistant community thread described CompreFace as having gone many months without an update and noted that it does not understand Frigate by itself. source
Checked against the maker's pages on 15 Oct 2026.