Frigate
App : A product you use directly
Local NVR for cameras with 100% local AI object detection.
The best local NVR for privacy-respecting video surveillance.
Security reflects where data goes, requested permissions and incidents on record at the verification date. It is not a guarantee of zero risk: check the deployment mode and the connections you enable.
What it does, concretely
Traditional video surveillance systems generate many false alerts because they rely on simple motion detection: wind in the trees or a shadow is enough to trigger a notification. Frigate solves this problem by applying AI-based object detection directly on your own hardware, without sending footage to a remote server.
In practice, Frigate is installed via Docker on a Linux server and analyses your IP camera streams in real time. It uses OpenCV and TensorFlow to identify specific objects such as people, cars, or animals, and only triggers a recording when a relevant object is detected. You can define precise zones within each camera's field of view to refine alerts.
Frigate integrates natively with Home Assistant via a dedicated component, and communicates with other automation platforms such as OpenHab or NodeRed through the MQTT protocol. It also offers RTSP re-streaming and low-latency live viewing via WebRTC.
This tool is aimed at people who want to manage their video surveillance independently, without relying on the cloud, and who are comfortable administering a Linux server and Docker. It is not suited to those looking for a plug-and-play solution with no technical configuration.
What it offers
- ·Real-time object detection (people, cars, animals) processed entirely locally using TensorFlow and OpenCV.
- ·Continuous 24/7 recording or recording triggered by object detection, with configurable retention rules.
- ·Definition of precise zones within each camera's field of view to trigger alerts only in areas of interest.
- ·RTSP re-streaming to limit the number of direct connections to cameras and reduce network load.
- ·Low-latency live viewing via WebRTC and MSE, with a multi-camera dashboard.
- ·Native integration with Home Assistant, as well as OpenHab, NodeRed, and any MQTT-compatible system.
- ·Custom detection models designed specifically for Frigate to further reduce false positives. (paid)
- ·Built-in mask and zone editor in the interface for visually configuring detection areas.
In practice
- Price
- Freemium
- Commercial use
- Commercial use: yes
- Available on
- web, linux, docker
- Licence
- MIT
- FR interface
- Yes
Security and data
- Rating
- Security: nothing to flag
- Your data
- Data stays local
- Known incidents
- Aucun.
- Maintenance
- Active
- Last release
- 15 October 2026
Links and repository checked on 25 October 2026 · added on 1 January 2020