Skip to main content

Camera Overview

Aegis works with cameras you already own. No proprietary hardware required.


Supported Camera Types

TypeExamplesHow to ConnectSetup Time
CloudBlink, RingLink your account with email + password2 minutes
LAN (ONVIF)DaHua, HikVision, Lorex, ReolinkAuto-discovered on your network1 minute
LAN (RTSP)Any IP cameraPaste the RTSP URL1 minute
WebcamBuilt-in laptop cam, USB camerasDetected automatically30 seconds
MobileiPhone, iPad, AndroidUse an IP Camera app5 minutes

How to Access Camera Settings

  1. Click the ⚙️ gear icon at the bottom of the sidebar
  2. Expand the Cameras section
  3. Choose:
    • Security Cameras → for Blink, Ring, and LAN cameras
    • Webcam (Builtin & USB) → for local devices and desktop capture

Which Should I Use?

Already have Blink or Ring cameras?Blink setup or Ring setup. Easiest path — just enter your account credentials.

Have IP cameras on your network?LAN Camera Discovery. Aegis auto-scans your network and finds them via ONVIF or the network scanner.

Just want to try it out?Use your webcam. Zero setup — select your device and go.

Have a spare phone?Use it as a camera. Install an IP camera app and point it anywhere.


How Cameras Feed the AI

Once a camera is added, here's what happens:

  1. Frames are captured — at the rate configured in Video Analysis settings
  2. The VLM analyzes them — the active Vision Model describes what it sees in each frame
  3. Clips are recorded — stored to your configured storage path with AI descriptions attached
  4. Alerts fire — if an event handler matches the description, notifications are sent
  5. The agent knows — the Aegis agent has full context of what every camera sees and can answer your questions

All cameras feed into the same pipeline. You can mix and match types — for example, two Blink cameras, one HikVision RTSP camera, and a USB webcam all running simultaneously.


Camera Limits

There's no hard limit on the number of cameras. Performance depends on your hardware:

SetupRecommended Cameras
CPU only, ≤8 GB RAM1–4 cameras
GPU (NVIDIA RTX), 16+ GB RAM4–12 cameras
Apple Silicon (M1–M4), 16+ GB4–12 cameras
Cloud VLM (OpenAI/Anthropic)Limited by API rate, not hardware

More cameras means more frames for the VLM to analyze. If inference falls behind, frames are automatically skipped to keep the system responsive.