Scale your intelligence,
not your headcount.
Whether you are piloting a 10-camera warehouse or securing a 5,000-camera smart city grid, Argu offers flexible edge and cloud deployment models.
Monthly
Tab Item Content
Tab Item Content
Annually
Save 20%
Pilot
$25
/cam/mo
- Up to 50 active cameras
- Secure Cloud processing (SaaS)
- 10 Custom NLP Rules
- Basic Webhooks & Slack integration
- Standard Support (48h SLA)
Most Popular
Scale
$35
/cam/mo
- Up to 250 active cameras
- Argu Edge Node included
- Unlimited NLP Rules
- Native VMS Integrations (Milestone, Genetec)
- Priority Support (12h SLA)
Enterprise
Custom
- Unlimited cameras
- Full Air-gapped on-prem deployment
- Fine-tuned custom vision models
- SIEM Log Forwarding (Splunk, Datadog)
- Dedicated Solutions Architect (1h SLA)
Why we win.
Breaking down the three pillars of Argu's architectural lead.
NLP Compilation
Traditional AI requires "training" or drawing tripwires. Argu uses a proprietary LLM-to-Vision compiler. You type a command, and we translate that into deterministic edge tracking parameters instantly.
Deterministic Edge
Cloud AI suffers from "Network Jitter." Argu runs on your server racks. If your internet goes down, the Edge Node continues to process, track, and log locally until connection is restored.
Native Sovereignty
We were built for "Compliance First" industries. Our architecture inherently satisfies GDPR and CCPA because we don't store or transmit raw PII-sensitive video frames to the public cloud.
Calculate your hardware footprint
See exactly how many Argu Edge Nodes you need based on your camera density.
Pricing FAQs
If you subscribe to the Scale tier with an annual contract, the required physical Argu Edge Nodes (1 node per 50 cameras) are shipped and licensed to you at no additional hardware cost. You simply pay the SaaS per-camera fee.
No. Access to our native VMS plugins (Milestone, Genetec, Avigilon) and our RESTful API is fully included in the Scale and Enterprise tiers. There are no hidden API payload limits or connection fees.
A rule is a single natural language prompt actively running against a camera stream. For example, “Alert if a person is holding a weapon” is one rule. If you run that rule across 10 cameras, it still counts as a single rule configuration. The Pilot tier limits you to 10 total configurations.