YunKan-TRT

YunKan-TRT (YunKan, "Sky View") is a self-hosted network video recorder (NVR) and video-management system (VMS) for home and small-business IP cameras that…

  • Home Automation
  • Media Applications
  • Security
  • Tools & Utilities
YunKan-TRT (YunKan, "Sky View") is a self-hosted network video recorder (NVR) and video-management system (VMS) for home and small-business IP cameras that runs on-device AI to detect motion, people, vehicles, faces, license plates (ALPR/LPR), pose and falls, packages, and baby cries, keeping all video on your local network with nothing uploaded to a vendor cloud. It is a privacy-first, single-container alternative to Frigate, Shinobi, ZoneMinder, Scrypted, Agent DVR, iSpy, and MotionEye, so you do not wire up Mosquitto, a database, a reverse proxy, or model downloads by hand. This TensorRT (TRT) build is the highest-throughput option for NVIDIA GPUs, recommended at 6 GB VRAM or more (8 GB or more for many high-resolution cameras). On first start it builds inference engines optimized for your exact GPU, which takes a few extra minutes once; the engines are cached on disk in the data directory and reused across restarts.

Platform availability

Available on 1 of 10 platforms

  • Unraid
  • TrueNAS (not listed)
  • Umbrel (not listed)
  • ZimaOS (not listed)
  • Proxmox (not listed)
  • Helm (not listed)
  • Mac (not listed)
  • Windows (not listed)
  • Linux (not listed)
  • Any server (not listed)

Health score?A 0-100 score computed nightly from four factors: maintained (40%: recent commits and steady activity), popular (25%: GitHub stars, log scale), easy to install (20%: platforms, install notes, setup guides, Docker image) and light to run (15%: minimum RAM, ARM support). When the source code has been checked, a fifth factor, security check (20%: automated scan for committed secrets, unsafe desktop-app settings, known-vulnerable dependencies and project hygiene; any serious finding caps it at 20), is added. Factors without data are left out and the rest are rescaled.

57/100 Fair

Maintained
100last commit 17 days ago
Popular
61 GitHub stars
Easy to install
351 install platform, install notes, Docker image
Light to run
No data

Based on 3 of 4 factors.

Availability: platform listings confirmed today, links checked 28 days ago

GitHub stars
1
Open issues
0
Last commit
2026-11-04
License
MIT
Activity
declining

Checked 13 days ago - source: GitHub (mrtian2016/yunkan-unraid-templates)

Resources & Compatibility

ARM (e.g. Raspberry Pi) is not supported.

Checked 54 days ago - source: Docker Hub image tags and published docs

First-install notes

  • First start & GPU Requires an NVIDIA GPU with at least 6 GB VRAM (8 GB or more for many high-resolution cameras). On first start YunKan-TRT builds TensorRT inference engines for your exact GPU, taking a few extra minutes once; the engines are cached on disk in the data directory and reused across restarts. source

Fetched 39 days ago - each note links to its source

Alternatives

No alternatives collected yet.

Common questions

Why does YunKan-TRT take longer on first start?

The TRT variant builds GPU-specific inference engines on your own GPU. This one-time build takes a few minutes, is cached under the data directory, and is reused across restarts and image updates.

Is YunKan charged per camera?

No, it is licensed per household rather than charged per camera.

Which camera connection methods does it support?

It supports RTSP, ONVIF, and GB28181 cameras.

Does it provide mobile or web apps?

It includes native Android and iOS apps, a Web client, and Home Assistant/HomeKit support.

Does camera footage leave my LAN or go to a cloud?

No, footage stays on your self-hosted device and nothing is uploaded to vendor servers.

Answers sourced from github.com, yun-kan.com

Community

Collected 53 days ago - public community threads