Products

Manage and analyse up to 32 high-resolution cameras with advanced AI that identifies faces, vehicles, and security threats across large-scale sites.

The Hikvision iDS-9632NXI-M8/X is a 32-channel DeepinMind network video recorder with eight SATA bays, RAID support and dual deep-learning engines. It runs facial recognition, perimeter protection, video structuralization and falling-object detection on the recorder itself, records cameras up to 32MP, and outputs up to 8K over HDMI. It has no built-in PoE and ships without hard drives.

At a glance
- 32 IP channels, up to 32MP; 320 Mbps in / 400 Mbps out; H.265+
- Recorder-side AI: facial recognition, perimeter protection, video structuralization, falling-object detection
- 8 SATA bays, up to 14TB each (112TB); RAID 0/1/5/6/10
- HDMI up to 8K (HDMI1) plus second HDMI and VGA outputs
- No built-in PoE (cameras via external switch); supplied with no HDD

The Hikvision iDS-9632NXI-P8 is a 32-channel DeepinMind Pro AI NVR with two on-board GPU engines for face recognition and natural-language search. It runs perimeter protection, people and vehicle differentiation and video structuralization on the recorder, stores footage across eight SATA bays with RAID, and outputs 8K over HDMI 1. It has no built-in PoE and ships without hard drives.

At a glance
- 32 IP channels, up to 32MP; DeepinMind Pro dual-GPU analytics
- Face recognition (up to 32 libraries / 100,000 images); perimeter, structuralization
- AcuSeek text search + AcuSearch image search; up to 600,000 targets/day
- 8 SATA bays, up to 20TB each, RAID 0/1/5/6/10
- 8K on HDMI 1; no built-in PoE; supplied with no HDD

The Hikvision iDS-9664NXI-M16/X is a 64-channel 3U DeepinMind network video recorder with sixteen SATA bays and dual deep-learning engines. It runs facial recognition, perimeter protection, video structuralization and falling-object (throwing-from-building) detection on the recorder, records cameras up to 32MP, and outputs up to 8K over HDMI. It has no built-in PoE and ships without hard drives.

At a glance
- 64 IP channels, up to 32MP; 400 Mbps in / 400 Mbps out; H.265+
- Recorder-side AI: facial recognition, perimeter protection, video structuralization, falling-object detection
- 16 SATA bays; RAID 0/1/5/6/10 for data protection
- HDMI1 up to 8K plus second HDMI and dual VGA outputs
- No built-in PoE (cameras via external switch); supplied with no HDD

The Hikvision iDS-9664NXI-M8/X is a 64-channel DeepinMind network video recorder with eight SATA bays and dual deep-learning engines. It runs facial recognition, perimeter protection, video structuralization and AcuSearch on the recorder itself, records cameras up to 32MP, and outputs up to 8K over HDMI. It has no built-in PoE and ships without hard drives.

At a glance
- 64 IP channels, up to 32MP; 400 Mbps in / 400 Mbps out; H.265+
- Recorder-side AI: facial recognition, perimeter protection, video structuralization, AcuSearch
- 8 SATA bays, up to 16TB per drive (128TB); RAID 0/1/5/6/10
- HDMI1 up to 8K plus second HDMI and dual VGA outputs
- No built-in PoE (cameras via external switch); supplied with no HDD

- Dual 4MP Lenses With 36× Optical Zoom: Two high-resolution lenses — a motorised 36× zoom and a wide-angle fixed lens — operate simultaneously to deliver long-range subject detail alongside broad scene awareness from a single unit, removing the trade-off between overview and close-up that limits single-lens PTZ cameras.
- On-Board ANPR, Face Recognition, and AI Behaviour Detection: AI analytics process locally without a backend server, covering automatic number plate recognition, face recognition, and behaviour-based detection for smoking, phone use, and missing PPE such as hard hats — critical for roadside, event, and construction compliance deployments.
- 4G/5G, Wi-Fi, GPS, and Bluetooth Connectivity: Multiple wireless options allow the unit to transmit live footage and alerts from locations without fixed network infrastructure. GPS tagging confirms deployment location for incident records, and Bluetooth supports local device configuration.
- 9-Hour Replaceable Battery With Magnetic/Tripod Deployment: A hot-swappable lithium-ion battery provides up to 9 hours of operation per charge. The magnetic base attaches to vehicle roofs in minutes; the tripod mount supports freestanding placement — neither requires drilling, cabling, or permanent infrastructure.

- On-Board ANPR Engine With >98% Accuracy: Built-in AI algorithm processes licence plate recognition locally on the camera at >99% capture rate and >98% LPR accuracy, eliminating reliance on a separate ANPR server for basic recognition tasks at single-lane entry points.
- Two Motorised Varifocal Lens Variants: The A variant ships with a 4–11mm lens suited to shorter-range or wider-lane installations (3–8m standoff), while the B variant offers an 8–32mm lens for longer distances or narrower lane framing. Motorised autofocus allows remote adjustment without physically accessing the housing.
- 140dB WDR for Entry Lane Performance: Entry and exit lanes are the most demanding WDR environments. The 140dB rating allows the sensor to retain plate detail in conditions where vehicle headlights, morning sun, and deep shadow occur simultaneously.
- Vehicle Classification and Colour Detection: Beyond plate reading, the ANPR engine classifies vehicles as car, van, bus, truck, or other, identifies body colour across 9 categories, and recognises 65 vehicle manufacturers — data logged alongside each capture event.

- 9MP ITS Imaging for High-Detail Plate Capture: The 9-megapixel resolution delivers the level of image detail required for vehicle and plate capture at checkpoint lanes, supporting accurate number plate reads and vehicle identification across a defined capture zone with greater pixel density than standard 4MP or 5MP ANPR units.
- ANPR and LPR Support With Vehicle Recognition: Automatic number plate recognition and licence plate recognition capability extends beyond simple plate capture, supporting vehicle classification and identification within a controlled lane or checkpoint environment for structured access and traffic evidence workflows.
- H.265 Compression for Multi-Lane Deployments: H.265 compression reduces the bandwidth and storage load of high-resolution ITS footage, helping sites with multiple checkpoint lanes or extended recording retention manage infrastructure costs in compatible recording systems.
- IR Illumination for 24-Hour Operation: IR illumination supports plate and vehicle capture during low-light and night-time conditions when the camera is correctly aligned to the capture zone, maintaining usable footage at checkpoints that operate around the clock.