Mantiswave Networks
Mantiswave AI-on-Edge Platform
AI/ML & Vision Processing on the Edge with Ultra Connectivity
Deploy powerful, real-time AI where it matters most. The Mantiswave AI/ML-on-Edge platform delivers high-performance AI/ML processing in a compact, 5G-connected embedded device, purpose-built for drones, robotics, industrial IoT, smart infrastructure, and research environments. Powered by a 26 TOPS NPU, it enables real-time inference and vision processing, supporting video analytics up to 4K resolution for applications such as surveillance, anomaly detection, object recognition, tracking, and many more. It supports USB, IP, and Smart Cameras as well as video file and image inputs, giving users complete flexibility in input data sources. With full support for custom model deployment, users can port and optimize their own datasets and AI models directly on the device.
For academia and R&D, the platform also serves as a hands-on AI/ML test-bed, enabling experimentation, prototyping, and learning in controlled and real-world environments on embedded systems. Connectivity is built-in with 5G, 4G LTE, Gigabit Ethernet, and Wi-Fi 5, ensuring low latency, high throughput, and secure data transfer across diverse applications.
With its modular design, user-friendly GUI, and rapid testing capabilities, Mantiswave AI-on-Edge provides a flexible, scalable solution for next-generation edge intelligence—spanning IoT, smart cities, drones, factories, security, classrooms, and advanced research labs.
AI/ML Processing on the EDGE
- Dual Core NPU @ 26 TOPS
- USB, IP and Smart Cameras
- 5G, 4G, Gigabit Ethernet, WiFi 5, BLE
- Support User Defined AI Model

- Comprehensive Dashboard
- upto 4K Video Processing
- Portable
- Cost Effective
Fit for all use cases
Defense Sector
Industry 4.0
Universities and R&D
Disaster Management
Underground Mining
Smart campus
Dashboard Features
- Configure: AI Model Flow, Input and output
- Deploy: Start/Stop inference, Model Porting
- Monitor: Inference Results, Output Videos, KPIs
- Manage: Model, Pre and Post Processing Methods
- Integrate: 3rd Party Apps
Dashboard Screenshots
Key Features

Edge AI Inference
Run real-time video analytics, object detection, segmentation and many more directly on the device.

Camera Integration
Flexible input with plug-and-play support for USB, IP and Smart cameras, video streams, and images.

Custom Model Integration
Deploy your own AI models, trained on custom datasets, and optimize them for real-world edge applications.

Vision Processing
Advanced edge vision analytics with recognition, tracking, and anomaly detection—scalable up to 4K video.

High Speed Connectivity
Seamless, data exchange via integrated 5G, Gigabit Ethernet and WiFi 5.

Compact & Modular
Fits into drones, robots, IIoT Gateways and embedded systems with ease.
Mantiswave Edge AI Platform Technical Specification
Main Processor | 64-bit Quad-core ARM Cortex-A76 @ 2.4GHz) |
RAM Options | 8/16 GB LPDDR4X |
AI Accelerator | Dual-core NPU (26 TOPS), supports TensorFlow Lite, ONNX, PyTorch |
Connectivity | 5G, 4G LTE, Wi-Fi 5, Bluetooth 5.0, Gigabit Ethernet |
5G Network | Fully compatible with Public 5G and 5G Private Network. |
Connectivity | |
I/O Ports | 2× micro HDMI, 2× USB 3.0, 2× USB 2.0, GPIO, CSI Camera |
Storage | Micro SD + USB/SSD via USB 3.0 |
Power Supply | 5V, 30 Watt |
AI Frameworks | TensorFlow, PyTorch, OpenCV etc. |
Form Factor | Modular & stack-able (90mm x 60mm x 60mm approx.) |
We endeavor to answer all inquiries within 24 hours on business days.
Email: contact@mantiswave.in