Lumana vs Rhombus
Nailing the basics is one thing.
Scaling the AI is another.
Rhombus nails the basics, but its AI is capped by what fits inside the camera. Lumana runs on NVIDIA GPUs with up to 100x more processing power, so your AI scales independently of your hardware — not the other way around.
A 30-minute walkthrough on the cameras you already own. No rip and replace.
Key differentiator
Lumana runs AI on NVIDIA GPUs, so your AI scales with what you need — not with what fits inside a camera.
Our technology
AI that scales past the camera
Rhombus's AI is capped by what fits inside the camera. Lumana runs on NVIDIA GPUs with up to 100x more processing power, so AI scales independently of your hardware.
Proven in the real world
At the largest technology conference in the US, Lumana turned 3.3 million person detections into just 89 alerts — reviewed in under 30 minutes total.
Every detection your team asked for
Because compute isn't trapped in the camera, you can run the detections you actually need — and add new ones over time, without waiting for the next generation of hardware.
A brain that perceives, reasons, and acts
A brain isn't one feature — it's three faculties working together. Here's what VIA-1 does on any IP camera you already own, from a single site to tens of thousands of cameras.
Perceive
See, with human-like perception
VIA-1 watches every feed in real time — object recognition, behavior analysis, and anomaly detection — and adapts to each camera's environment, getting more accurate the longer it runs.
People, vehicle, and object detection on any IP camera
Continuous situational learning, per camera
Behavior and anomaly detection, not just motion
reason
Understand, and find anything fast
Ask in plain language. VIA-1 reasons across every camera in your organization, turning hours of footage into the exact moment you need — in seconds.
Free-text search: "person in a black shirt walking through a door"
Multi-object search across 40+ attributes
Occupancy and operational insights over time
act3
Automated, user-defined responses
When something that matters happens, VIA-1 doesn't just flag it — it acts. Define the response once in plain language, and it runs on its own, every time.
Custom alerts written in natural language
Notifications, audible messages, alarms, GPIO, API, and webhooks
If AI capacity weren't a hardware question, what would you want from your video security?
See how Lumana makes that real — AI on NVIDIA GPUs, running on the cameras you already own, with new models added in software.