Deepsight Face

Deepsight Face is a face recognition solution based on Deep Learning. It is designed to enable offline inference and, thus, eliminating the need for expensive Cloud based API subscriptions. Deepsight runs on the machine locally as a http service. Developers can build their custom face solutions around this and deploy it on machines that can run offline anywhere.


Deepsight Face is currently supported in Linux and Windows Operating Systems on x64 platform.

It is available with or without GPU Acceleration. The free version comes without it.


  • No need for internet. 100% offline inference.

  • REST API. Easy. Simple.

  • Built in C++. Optimized to leverage powerful CPU extensions like SSE 2,4,4.1 as well as multi-core/thead architectures

  • Available with or without GPU (CUDA) Acceleration.

  • Plugin architecture - all neural nets are individual plugins and can be updated/replaced/removed.

  • Periodic updates to individual nets with improved accuracy.

  • Easy setup. Run the installer and launch the app!

See the getting started guide for instructions on how to get it up and running.