DS-7608NXI-K2/VPro is an advanced 8-channel IP recorder from Hikvision’s K/VPro series, designed for demanding IP CCTV systems where the effectiveness of analysis and the speed of searching for objects meeting user-defined criteria are of paramount importance. The recorder can be connected to IP cameras with a resolution of up to 12 Mpix, ensuring the level of detail required for identification. Independent HDMI (up to 4K) / VGA (up to 1080p) outputs allow two monitors to be connected, each with a separately configurable camera view. The available input bandwidth of 80 Mb/s and output bandwidth of 160 Mb/s ensure smooth recording and video decoding, even under heavy system load. The recorder can be fitted with 2 hard drives equipped with SATA interfaces, each with a capacity of up to 16 TB. When the storage is full, automatic overwrite is activated, protecting the most recent data from loss. The USB ports (2 x USB 2.0) enable the rapid export of recordings to external storage media. A Gigabit network port (1 × RJ‑45 10/100/1000 Mb/s) enables communication with IP cameras and integration with the local network and the internet.The K/VPro series devices utilise Hikvision’s advanced image analysis algorithms, including AcuSense, AcuSeek, AcuSearch and Motion Detection 2.0. AcuSense filters out false alarms by classifying objects as either a person or a vehicle, significantly enhancing the effectiveness of perimeter security. The AcuSeek and AcuSearch functions allow for quick searching of recordings based on AI metadata generated by the cameras, such as clothing colour or object type, which significantly speeds up the process of finding events of interest. The recorder also supports facial recognition, enabling the capture, comparison and search of facial images within 16 libraries with a total capacity of up to 50,000 images. The system works with ANPR cameras for number plate recognition, as well as PTZ and VCA cameras, enabling advanced monitoring scenarios such as line crossing or intrusion. The AI metadata-based search available in the AcuSeek/AcuSearch functions supports only the English language.PLEASE NOTE! The latest version of the software is available at this linkKey features:Supports up to 8 IP cameras from Hikvision or other manufacturers via ONVIFRecording at resolutions of up to 12 MpixIndependent HDMI / VGA outputsmaximum input/output bandwidth – 80 / 160 Mb/ssupports 2 SATA hard drives with a capacity of up to 16 TB eachAI functions provided by the recorder: perimeter protection (2 channels) or face detection and analysis (2 channels) or Motion Detection 2.0 (8 channels)16 face image libraries (max. 50,000 images)support on all camera channels for AcuSense and Motion Detection 2.0support for cameras with built-in VCA analytics (virtual line, intrusion detection, etc.) AcuSearch / AcuSeek – up to 8 channels (feature available on the recorder when cameras generating the relevant AI metadata are connected) integration with the Guanlan platform ANPR (number plate recognition) – only if Hikvision IP cameras have the ANPR functionsupported compression: H.265 (+) / H.264 (+)features supported: ANRsupports up to 4 IP speakersaudio I/O for two-way communicationAlarm I/O: 4 / 12 x USB 2.0Ethernet ports: 1 x RJ-45 10/100/1000 Mb/sdesktop app iVMS-4200 and smartphone app Hik-Connect (Android, iOS)cloud access P2PAcuSeek – a revolution in video search for Hikvision recordersAcuSeek is a groundbreaking feature available in the latest Hikvision IP recorders, which completely transforms the way video footage is searched. Using advanced next-generation artificial intelligence, the system can understand both what the camera sees and what the user types, combining image and language analysis into a single, cohesive environment. This solution is like a personal assistant who knows all the recordings by heart, understands the user’s intentions and instantly pinpoints exactly what they are looking for without the need to sift through hours of video footage. Whether you need to find a specific person, vehicle or incident, AcuSeek delivers precise results in just a few seconds. This technology is based on a large multimodal AI model called Guanlan, which analyses data in layers:Image recognition: identifies people, vehicles, objects, animals, as well as their behaviours and interactionsUnderstands text: processes natural user queries, e.g. “a person with a rucksack entering a building”, without the need for technical languageConnects context: intelligently matches the description to the relevant parts of the recording, taking into account time, place and situationThanks to this multi-layered analysis, AcuSeek not only ‘sees’ and ‘reads’, but above all understands, making it one of the most advanced video search tools available on the market. This feature significantly boosts the efficiency of surveillance operators, security services and private users alike, offering intuitive operation, instant results and unrivalle