Shinobi

Not yet benchmarked

Vendor: Shinobi Systems
License: GPLv3/AGPLv3 (CE) ; EULA propriétaire (Pro)
Platforms: Linux, Docker, Windows via Docker/WSL

Lightweight modern Node.js VMS supporting thousands of cameras, easy Docker deployment and a plugin ecosystem. Not yet benchmarked on our test bench.

Strengths

Lightweight Node.js core, easy multi-arch Docker deployment (incl. ARM/Raspberry Pi).
Broad camera support and a plugin ecosystem (object detection built in; face via plugin).

Limitations

Facial recognition is not in the core — it comes via a plugin (DeepStack).
Confusing CE-vs-Pro split with a non-standard commercial EULA on Pro.
No clear versioned stable releases (rolling updates); no encryption at rest.

Criteria

Self-hosted (runs on your hardware)	✓
Works without a mandatory vendor account	✓
Works with any ONVIF/RTSP camera (no hardware lock-in)	✓
Built-in facial recognition	✗
Encrypts stored footage at rest	✗
Remote access without a vendor cloud	✓
Web + desktop + mobile apps	✗

Sources

https://github.com/ShinobiCCTV/Shinobi/blob/master/LICENSE — officiel (licence CE) (2026-07-21)
https://docs.shinobi.video/detect/deepstack — officiel (reco faciale = plugin DeepStack) (2026-07-21)
https://shinobi.video/pro — officiel (édition Pro) (2026-07-21)

This product has not yet been tested on our bench. The information above comes from dated documentary sources; no comparative verdict is published until a real test has taken place.