Jetson Orin NX vs AGX Orin: Specs, Power, and Which to Buy

The short version: Jetson Orin NX is the compact, lower-power module built for robots, drones, and embedded vision systems, while Jetson AGX Orin is the larger, more powerful module for machines that need to run several demanding AI pipelines at once. Both belong to NVIDIA's Orin family and share the same Ampere GPU architecture and Arm Cortex-A78AE CPU design, but they differ sharply in raw AI performance, memory bandwidth, power envelope, and physical form factor. This guide compares the verified specs side by side, explains what each difference actually means for a real design, and points to the developer kits and carrier boards each one needs.

Jetson Orin NX vs AGX Orin: the short answer

Jetson Orin NX ships in 8GB and 16GB versions, rated at up to 117 and 157 TOPS of AI performance respectively, in a small 69.6 x 45 mm SO-DIMM module that draws between 10 W and 40 W. Jetson AGX Orin ships in 32GB and 64GB versions, rated at up to 241 and 275 TOPS, in a larger 100 x 87 mm module that draws between 15 W and 60 W, according to NVIDIA's own Jetson Orin technical specifications page. The gap is not just a TOPS number: AGX Orin also carries up to twice the CPU cores, twice the memory bus width, and double the memory bandwidth of Orin NX, which matters more than the headline AI score for many workloads.

If your project is smaller than either of these — a single camera, a light object-detection model, tight power and cost budgets — the Jetson Orin Nano (up to 67 TOPS, 7-25 W) is worth reading about before you commit to either module covered here.

Specs compared: performance, memory, and power

These figures come from NVIDIA's published Jetson Orin technical specifications, which list current "Super" performance ratings after NVIDIA's 2024 software update raised the ceiling on Orin NX and Orin Nano through higher power and clock limits.

Two rows explain most of the real-world gap. First, memory bandwidth: AGX Orin's 256-bit bus moves data twice as fast as Orin NX's 128-bit bus, which matters when several models or high-resolution camera streams are reading and writing memory concurrently. Second, DLA count: Orin NX 8GB gets a single dedicated deep learning accelerator, while every other module here — including Orin NX 16GB — gets two, freeing the GPU to run additional models in parallel instead of queuing behind the same engine.

Jetson Orin NX: the compact module for embedded designs

Jetson Orin NX targets products where board space and power budget are fixed early in the design: mobile robots, drones, handheld scanners, and compact industrial vision systems. NVIDIA's Jetson Modules page describes it as delivering up to 5x the performance and twice the CUDA cores of the previous-generation Jetson Xavier NX, in what it calls the smallest Jetson form factor. The 16GB variant's extra CPU cores, second DLA, and doubled memory over the 8GB variant make it the better pick whenever more than one model needs to run at once — for example, an object detector feeding a tracking or segmentation stage rather than a single classifier running alone.

Because Orin NX is a SO-DIMM module rather than a complete computer, it needs a carrier board. Real Amazon listings for Orin NX kits — including 16GB units bundled with a power supply, wireless card, and enclosure — reflect the current 157 TOPS "Super" rating, confirming that figure is the one shipping in current hardware rather than a launch-day number that has since changed:

Disclosure: this article contains affiliate links. If you buy through them we may earn a small commission at no extra cost to you.

Yahboom Jetson Orin NX 16GB 157TOPS Development Kit for AI Edge, with 48W Power Supply, Wireless Network Card, Enclosure

• 【Core Parameters】★AI Perf: 117/157 TOPS★GPU: 1024-core N-VI-DIA Ampere architecture GPU with 32 Tensor Cores★CPU: 8-core Arm Cortex-A78AE v8.2 64-bit CPU 2MB L2 + 4MB L3★Memory: 16GB 128-bit LPDDR5 | 102.4GB/s★Storage: Supports external NVMe 【Note: This kit does not include a SSD and pre-installed system. User need to provide your own NVMe M.2 SSD of at least 256GB and flash the operating system onto it yourself. 】
• 【Empowered by Large Al Model, Enhanced Human-Computer Interaction】Jetson Orin Super leverages three AI models and incorporates an AI voice interaction module. This multimodal visual system matches the scene being described, enabling environmental awareness and AI visual gameplay. Combined with a large-scale voice module and camera, it enables speech-to-text, semantic analysis, natural conversation, and real-time video analysis, enabling advanced embodied AI applications.
• 【Revolutionize the Industry】Jetson Orin NX modules deliver unmatched performance and efficiency for small, low-power robotics and autonomous machines, making them ideal for drones, handheld devices, and more. The module can be easily used in advanced applications in manufacturing, logistics, retail, agriculture, medical and life sciences, and comes in a highly compact and energy-efficient package.
• 【Revolutionizing AI with Unmatched Performance】The Jetson Orin NX system module adopts the Ampere architecture GPU, a new generation of deep learning and vision accelerators, high-speed I/O, and fast memory bandwidth to support multiple AI application processes. Granular structured sparsity to improve the operating throughput of Tensor Core, and can use larger and more complex AI model development solutions in natural language understanding, 3D perception and multi-sensor fusion.
• 【Tutorial materials provided】The JETSON system based on Ubuntu 22.04 provides a complete desktop Linux environment with accelerated graphics, supporting NVIDI-ACUDA 12.6, TensorRT 10.7.0, cuDNN 9.6.0, OpenCV 4.10.0, etc. The performance on AI LLM, VLM and visual Transformer is significantly improved compared with the previous generation.

$1,329.99

Buy on Amazon

When shopping for any Orin NX kit, check three things before the price: which module size it uses (16GB for multi-model workloads, 8GB for a single lighter model), whether the listed TOPS rating matches the current "Super" figures above rather than an outdated launch-day number, and whether the carrier board, power supply, and cooling are included — Orin NX modules are sold bare far more often than AGX Orin, so missing accessories are the most common first-build mistake.

Jetson AGX Orin: the flagship for demanding pipelines

AGX Orin is the module NVIDIA positions for autonomous machines that need to process multiple concurrent AI pipelines from multiple sensors — the kind of workload that shows up in manufacturing robots, delivery and logistics platforms, and higher-tier medical and retail devices. Its 12-core CPU (versus Orin NX's 6 or 8 cores) and doubled memory bandwidth matter as much as its TOPS ceiling once a pipeline stops being "run one model" and becomes "run perception, planning, and a vision-language model together, fed by several cameras." NVIDIA also sells an Industrial variant of the 64GB module rated at 248 TOPS with a wider 15-75 W power range for extended-temperature deployments — a niche option outside the standard 32GB/64GB choice covered above.

Because every AGX Orin module ships as a 100 x 87 mm board with an integrated thermal transfer plate and 64GB of onboard eMMC storage, it behaves closer to a small, complete computer than Orin NX does, which is reflected in the developer kits sold around it:

Waveshare Jetson AGX Orin Developer Kit, Server-Class AI Performance At The Edge, Up to 275 Tops 64GB Memory

• Provide online user manual, please check the manual carefully before using
• The NV Jetson AGX Orin Developer Kit includes a high-performance, power-efficient Jetson AGX Orin module with options for 32GB/64GB memory, up to 275 TOPS and 8X the performance of the last generation for multiple concurrent AI inference pipelines, for running the NV AI software stack.
• This developer kit lets you create advanced robotics and edge AI applications for manufacturing, logistics, retail, service, agriculture, smart city, healthcare, and life sciences.
• The Jetson AGX Orin provides 8X the performance of Jetson AGX Xavier with the same compact form factor and compatible pinouts, integrating NV Ampere architecture GPU, Arm Cortex-A78AE CPU, next-generation deep learning and vision accelerator.
• High-speed interface, faster memory bandwidth, and multi-mode sensor support, for supporting multiple concurrent AI application channels.

$5,843.99

Buy on Amazon

For teams building around multiple sensor types rather than a single camera — combining vision with radar, LiDAR, or IMU data before the AI stage — AGX Orin's extra memory bandwidth is typically the more useful upgrade over Orin NX than the raw TOPS number, and it's worth reading our guide to sensor fusion before locking in a module for that kind of design.

Form factor and power: why this choice locks in your board design

Orin NX and Orin Nano share the exact same 69.6 x 45 mm, 260-pin SO-DIMM form factor and mounting, which is why NVIDIA's own Jetson Orin Nano Super Developer Kit ships with "a reference carrier board compatible with all Orin NX and Orin Nano modules," per NVIDIA's Jetson Orin product page. In practice that means a team can prototype on the lower-power Orin Nano module and later swap in an Orin NX module on the same carrier board design as performance needs grow — without a board respin.

AGX Orin does not share that socket. It uses a much larger 100 x 87 mm module with a 699-pin Molex Mirror Mezz connector, so a design built around AGX Orin cannot drop down to an Orin NX or Orin Nano module later without redesigning the carrier board. That single mechanical fact is often a bigger factor in the NX-vs-AGX decision than the performance numbers: pick AGX Orin only when the project's compute needs are already known to require it, or when the extra board space and power budget genuinely aren't constraints.

Developer kits: how to actually get started

NVIDIA sells a dedicated Jetson AGX Orin Developer Kit — a complete 110 x 110 x 71.65 mm system that, per NVIDIA, "shares one System-on-Chip architecture" with every Orin module and "can emulate any of the Jetson Orin modules," which makes it a reasonable starting point even for software aimed at AGX Orin 32GB or a production carrier board. There is no equivalent standalone "Orin NX developer kit" from NVIDIA; instead, the smaller Jetson Orin Nano Super Developer Kit's carrier board — confirmed above to accept Orin NX modules — is the closest official starting point, alongside third-party carrier boards sold by companies that build for the Orin NX/Nano socket specifically. Seeed Studio, Waveshare, and Yahboom are among the vendors currently selling complete Orin NX carrier-board kits — bundling the module, board, power supply, and enclosure into one purchase — which is the practical reason most teams buy Orin NX as a kit rather than sourcing a bare module and carrier board separately.

One planning advantage of buying either module through NVIDIA's own developer-kit path rather than a bare module: NVIDIA describes the Jetson family as running on "unified NVIDIA CUDA-X software," so code, containers, and models built against the JetPack SDK on an Orin NX carrier board generally carry over to an AGX Orin design later — and vice versa — without a software rewrite, even though the two modules sit at very different points on the performance and power scale covered above.

If your project doesn't need on-device inference at all yet — you're still training or fine-tuning the model that will eventually run on Orin NX or AGX Orin — that workload belongs on a desktop or datacenter GPU rather than a Jetson module; see our comparison of cloud GPU providers for options that don't require buying hardware up front.

Which one should you buy?

Buy Orin NX (8GB or 16GB) when the product is a robot, drone, or embedded vision system with a fixed, tight power and space budget, when it needs to plug into the same carrier board family as Orin Nano across a product line, or when the workload is one to two concurrent models. Buy AGX Orin (32GB or 64GB) when the machine runs several demanding models simultaneously, ingests high-bandwidth sensor data from multiple sources, needs onboard eMMC storage without an external SSD, or when the 75 W Industrial variant's extended temperature range is a requirement. If you're not yet sure the project needs a Jetson at all, compare against a Raspberry Pi accelerator first in our Raspberry Pi AI HAT+ vs Jetson guide — many single-camera projects don't need Orin-class compute.

Orin NX kits currently listed on Amazon for this keyword, pulled automatically:

Bestseller No. 1

Yahboom Jetson Orin NX 16GB 157TOPS Development Kit for AI Edge, with 48W Power Supply, Wireless Network Card, Enclosure

• 【Core Parameters】★AI Perf: 117/157 TOPS★GPU: 1024-core N-VI-DIA Ampere architecture GPU with 32 Tensor Cores★CPU: 8-core Arm Cortex-A78AE v8.2 64-bit CPU 2MB L2 + 4MB L3★Memory: 16GB 128-bit LPDDR5 | 102.4GB/s★Storage: Supports external NVMe 【Note: This kit does not include a SSD and pre-installed system. User need to provide your own NVMe M.2 SSD of at least 256GB and flash the operating system onto it yourself. 】
• 【Empowered by Large Al Model, Enhanced Human-Computer Interaction】Jetson Orin Super leverages three AI models and incorporates an AI voice interaction module. This multimodal visual system matches the scene being described, enabling environmental awareness and AI visual gameplay. Combined with a large-scale voice module and camera, it enables speech-to-text, semantic analysis, natural conversation, and real-time video analysis, enabling advanced embodied AI applications.
• 【Revolutionize the Industry】Jetson Orin NX modules deliver unmatched performance and efficiency for small, low-power robotics and autonomous machines, making them ideal for drones, handheld devices, and more. The module can be easily used in advanced applications in manufacturing, logistics, retail, agriculture, medical and life sciences, and comes in a highly compact and energy-efficient package.
• 【Revolutionizing AI with Unmatched Performance】The Jetson Orin NX system module adopts the Ampere architecture GPU, a new generation of deep learning and vision accelerators, high-speed I/O, and fast memory bandwidth to support multiple AI application processes. Granular structured sparsity to improve the operating throughput of Tensor Core, and can use larger and more complex AI model development solutions in natural language understanding, 3D perception and multi-sensor fusion.
• 【Tutorial materials provided】The JETSON system based on Ubuntu 22.04 provides a complete desktop Linux environment with accelerated graphics, supporting NVIDI-ACUDA 12.6, TensorRT 10.7.0, cuDNN 9.6.0, OpenCV 4.10.0, etc. The performance on AI LLM, VLM and visual Transformer is significantly improved compared with the previous generation.

$1,329.99

Buy on Amazon

Bestseller No. 2

seeed studio NVIDIA Jetson Orin NX 16GB Edge AI Device - reComputer J4012, 4xUSB 3.2, M.2 Key E & Key M Slot, Pre-Installed Jetpack System with NVIDIA Jetpack on 128GB NVMe SSD

• 【Brilliant AI Performance for production】 on-device processing with up to 100 TOPS AI performance with low power and low latency, Due to the high thermal demands of Super mode, only the J30 Series supports upgrading to Super mode via the JetPack 6.2 update
• 【Hand-size edge AI device】 compact size at 130mm x120mm x 58.5mm, includes NVIDIA Jetson Orin NX 16GB production module, a cooling fan with a heatsink, enclosure, and a power adapter. Support desktop, wall mount, fit in anywhere
• 【Expandable with rich I/Os】4x USB 3.2, HDMI 2.1, 2xCSI, 1xRJ45 for GbE, M.2 Key E, M.2 Key M, CAN, and GPIO
• 【Accelerate solution to market】pre-installed Jetpack with NVIDIA JetPack 5.1 on the included 128GB NVMe SSD, Linux OS BSP, 128GB SSD, support Jetson software and leading AI frameworks and software platforms
• 【Comprehensive certificates】FCC, CE, RoHS, UKCA

$1,499.00

Buy on Amazon

Bestseller No. 3

Yahboom Jetson Orin NX 16GB 157TOP Super Kit, Embedded and Edge Systems for AI Large Model 16GB Core Module Mini PC Kit

• 【Core Parameters】★AI Perf: 117/157 TOPS★GPU: 1024-core N-VI-DIA Ampere architecture GPU with 32 Tensor Cores★CPU: 8-core Arm Cortex-A78AE v8.2 64-bit CPU 2MB L2 + 4MB L3★Memory: 16GB 128-bit LPDDR5 | 102.4GB/s★Storage: Supports external NVMe.
• 【Empowered by Large Al Model, Enhanced Human-Computer Interaction】Jetson Orin Super leverages three AI models and incorporates an AI voice interaction module. This multimodal visual system matches the scene being described, enabling environmental awareness and AI visual gameplay. Combined with a large-scale voice module and camera, it enables speech-to-text, semantic analysis, natural conversation, and real-time video analysis, enabling advanced embodied AI applications.
• 【Revolutionize the Industry】Jetson Orin NX modules deliver unmatched performance and efficiency for small, low-power robotics and autonomous machines, making them ideal for drones, handheld devices, and more. The module can be easily used in advanced applications in manufacturing, logistics, retail, agriculture, medical and life sciences, and comes in a highly compact and energy-efficient package.
• 【Revolutionizing AI with Unmatched Performance】The Jetson Orin NX system module adopts the Ampere architecture GPU, a new generation of deep learning and vision accelerators, high-speed I/O, and fast memory bandwidth to support multiple AI application processes. Granular structured sparsity to improve the operating throughput of Tensor Core, and can use larger and more complex AI model development solutions in natural language understanding, 3D perception and multi-sensor fusion.
• 【Tutorial materials provided】The JETSON system based on Ubuntu 22.04 provides a complete desktop Linux environment with accelerated graphics, supporting NVIDI-ACUDA 12.6, TensorRT 10.7.0, cuDNN 9.6.0, OpenCV 4.10.0, etc. The performance on AI LLM, VLM and visual Transformer is significantly improved compared with the previous generation.

$1,448.99

Buy on Amazon

Frequently asked questions about Jetson Orin NX and AGX Orin

What is the difference between Jetson Orin NX and Jetson AGX Orin?

Orin NX is the smaller, lower-power module (8GB or 16GB, 10-40 W, up to 157 TOPS) built for embedded and mobile designs, while AGX Orin is the larger, higher-power module (32GB or 64GB, 15-60 W, up to 275 TOPS) built for machines running multiple concurrent AI pipelines. AGX Orin also has up to twice the memory bandwidth and up to twice the CPU cores of Orin NX, and uses a different, larger physical connector.

What is the Jetson AGX Orin used for?

NVIDIA positions AGX Orin as an AI computer for energy-efficient autonomous machines running multiple concurrent inference pipelines, citing applications spanning manufacturing, logistics, retail, and healthcare, alongside generative AI, robotics, and computer vision workloads more broadly.

What is the difference between Jetson Orin Nano and Jetson Orin NX?

Both use the same 69.6 x 45 mm SO-DIMM module socket, but Orin Nano tops out at 67 TOPS on a 7-25 W power budget, while Orin NX reaches up to 157 TOPS on a 10-40 W budget with more CPU cores and, on the 16GB variant, a second DLA. See our dedicated Jetson Orin Nano guide for the full breakdown of that module.

What are the different Jetson Orin models?

The Orin family covers four module tiers: Orin Nano (4GB and 8GB, up to 67 TOPS), Orin NX (8GB and 16GB, up to 157 TOPS), and AGX Orin (32GB and 64GB, up to 275 TOPS, plus a 64GB Industrial variant rated at 248 TOPS with a wider temperature and power range).

Can I use the same developer kit for Orin NX and AGX Orin?

No. AGX Orin has its own dedicated developer kit built around its larger module and connector. Orin NX has no standalone NVIDIA developer kit; it shares a carrier-board socket with Orin Nano, and NVIDIA's Jetson Orin Nano Super Developer Kit carrier board is explicitly built to accept Orin NX modules as well.

Do Orin NX and AGX Orin run the same software?

Yes, at the platform level. NVIDIA states that the Jetson family "uses unified NVIDIA CUDA-X software," and the same JetPack SDK targets both modules, so code written for one generally ports to the other — though the amount of headroom for larger models or more concurrent pipelines still depends on the memory and compute differences covered above.

For the module below both of these in size and power, see our Jetson Orin Nano guide, or step back further with our Raspberry Pi AI HAT+ vs Jetson comparison if you're still deciding whether you need a Jetson at all.

Last update 2026-10-01. Price and product availability may change.