Skip to product information
Jetson Orin Nano Field Kit
The "Day 1" Edge AI Development Platform. Zero Setup. 100% Ready.
Stop spending days configuring drivers, compiling CUDA libraries, and debugging Docker containers. The Jetson Orin Nano Field Kit is the only development platform that arrives fully configured with a production-grade AI software stack on a bootable NVMe SSD.
It’s not just hardware; it is a complete, offline-capable application layer built on top of the powerful Jetson Orin Nano Super Developer Kit, designed for engineers who need to demonstrate value immediately.
This kit transforms the standard developer board into a ruggedized, portable workstation:
We've done the heavy lifting so you don't have to. Every unit ships with JetPack 6.2 and our open-source application layer:
👁️ Computer Vision Stack
🧠 Generative AI & LLMs
🛠️ Developer Tools
Get the kit that respects your time. Build the future, don't just configure it.
The Jetson Orin Nano Field Kit (Roboflow Edition)
Sale price
$799.00
Regular price
$900.00
Jetson Orin Nano Field Kit
Why This Kit?
- 0 to AI in Minutes: Boots directly into a custom, optimized environment.
- Field-Ready Networking: Automatically creates its own WiFi Hotspot - connect directly to the device in the field without external routers.
- Production Software Stack: Comes with MediaMTX, Roboflow Inference, Ollama, Llama.CPP, Open WebUI pre-installed, a few LLM and vision models, GPU-accelerated libraries, etc.
- Offline Capable: Runs LLMs, Object Detection, and Voice Assistants entirely locally.
Included Hardware
- NVIDIA Jetson Orin Nano Super Developer Kit: 1024-core NVIDIA Ampere architecture GPU with 40 TOPS of AI performance.
- 256GB NVMe SSD (Bootable): Pre-flashed with our custom "Field Kit" system image. High-speed storage drastically improves system responsiveness compared to microSD cards.
- Dual IMX219 Stereo Cameras: 8MP Sony sensors pre-mounted and calibrated for depth sensing and stereo vision.
- USB WiFi Adapter: Dedicated chipset (8821cu) for stable AP/Hotspot mode.
- Custom 3D-Printed Field Case: Protects the electronics and provides a rigid adjustable mounting for dual imx219 camera
The Software Advantage (Pre-Installed)
- Roboflow Inference Server: GPU-accelerated Docker container with TensorRT optimization.
- MediaMTX Streaming: Ultra low-latency RTSP/WebRTC/HLS video streaming server.
- Instant Object Detection: Runs YOLOv8, YOLOv11, and RF-DETR out of the box.
- Browser-Based HUD: View real-time inference overlays directly in Chrome/Safari via http://box.local.
- Local LLM Runtime: Open WebUI with Ollama backend.
- Pre-Loaded Models: Optimized Qwen2.5 and Ministral-3B models running locally.
- Voice Assistant: Wake-word enabled assistant setup with LiveKit integration.
- Claude Code & VS Code Server: Edit code directly on the device.
- Kiwix Server: Offline Wikipedia and technical documentation for remote work.
- Docker + NVIDIA Runtime: Pre-configured container runtime.
Perfect For
- Rapid Prototyping: Go from unboxing to live demo in under 10 minutes.
- Field Technicians: Carry a self-contained AI unit that creates its own network.
- Robotics & Drones: Low-latency stereo vision and depth sensing ready to go.
- Offline Agents: Build privacy-focused voice and chat agents that don't touch the cloud.
Technical Specifications
- GPU: NVIDIA Ampere architecture with 1024 CUDA cores and 32 Tensor cores.
- CPU: 6-core Arm® Cortex®-A78AE v8.2 64-bit CPU.
- Memory: 8GB 128-bit LPDDR5.
- Storage: 256GB NVMe SSD (PCIe Gen3 x4).
- Connectivity: Gigabit Ethernet, 802.11ac WiFi (Hotspot capable).
- Power: DC Jack (Power adapter included).