Our Cloud plans (Production & Enterprise) run entirely on EyePop.ai’s managed infrastructure, allowing you to get started immediately without managing hardware. Our On-Premise plan is designed for localized surveillance platform integrations, running directly on your edge devices to ensure minimal latency and keep data completely within your local network.

Regardless of the tier you choose, data exclusivity and privacy are completely guaranteed. On the Production plan, your data is fully isolated from other tenants. On the Enterprise and On-Premise tiers, you receive enterprise-grade security controls alongside custom compliance support tailored to your organization’s internal standards.

Every account includes close collaboration with our account engineering team for architecture review, customer strategy around new use cases, and deployment support. For advanced needs down the road, additional data labeling, model training, or dedicated ML engineer hours are available on a transparent, scoped-work basis.

Absolutely. You can use a standard Cloud account to prototype and test your logic against our cloud endpoints immediately. Once you're ready to test on-premises hardware, we provide a free 3-month license for one internal lab box so your team can thoroughly test in a staging environment.

A compute unit measures the processing resources your AI models use. The Production plan includes 4,000 compute units per month. If you exceed this, additional units are billed seamlessly at $0.05 per unit so your application never experiences downtime. The Enterprise plan includes dedicated servers with unlimited usage and zero overage fees.

Production plans use "Auto wake" servers, which spin up automatically when a request is made—ideal for general use and keeping costs low. Enterprise plans feature dedicated, "Always-up" servers that remain running 24/7 to guarantee ultra-low latency and instant response times for mission-critical apps.

A "box" refers to an individual physical edge device or localized server running EyePop.ai on-site. Our software licensing scales down on a per-device basis as your total volume of connected boxes increases.

Yes, you will purchase your own hardware directly. To ensure a smooth rollout, our team will guide you through the exact technical specifications needed. The estimated hardware cost is roughly $1,500 per box depending on your specific requirements.

This one-time fee covers your end-to-end system initialization. Our engineers will handle the upfront network configuration, seamless integration with your existing surveillance platform, and custom model tuning to ensure high-accuracy performance right out of the gate.

Our edge-native architecture ensures your video streams never leave the local network. EyePop.ai runs processing and inference directly on your on-premise hardware, transmitting only lightweight, privacy-compliant metadata back to the system.