GPUs & Accelerators

Google Coral USB Accelerator

A compact USB Edge TPU accelerator for low-power TensorFlow Lite inference on existing Linux, macOS, Windows, and Raspberry Pi systems.

Vendor

Google Coral

Price

$59.99

Starts at

AgricultureManufacturingRetailOther
Google Coral USB Accelerator connected by USB-C

Quick Summary

Why it matters

The Coral USB Accelerator is the opposite end of the AI hardware spectrum from H100-class GPUs. It gives edge builders a cheap, low-power way to add dedicated INT8 inference to cameras, kiosks, prototypes, and small embedded systems.

Specs

Simplified decision specs

ML Accelerator

Google Edge TPU coprocessor

AI Performance

4 TOPS INT8 / 2 TOPS per watt

Edge Profile

USB 3.0 Type-C / 65 mm x 30 mm / TensorFlow Lite

Best For

Where this product fits best

Low-power edge inference
Raspberry Pi and small Linux prototypes
TensorFlow Lite computer vision workloads

Used In

Example deployment contexts

Smart camera prototypes
Retail and kiosk inference
Agriculture and industrial edge sensors

Black Scarab POV

Where we see the real value

This is a useful catalog anchor for tiny edge AI. It will not run general LLM workloads, but it is excellent for cost-sensitive INT8 TensorFlow Lite models where low power, simple USB setup, and local inference matter.

Category

GPUs & Accelerators

Power Profile

Low (<25W)

Deployment Types

Indoor, Mobile, Industrial

Vendor

Google Coral

Price

$59.99

Starts at

Source

Official product page

Coral lists the USB Accelerator starting at $59.99 with multiple distributor links, while noting that some Coral products may face stock delays. Verify live stock before adding to a deployment basket.

Link Strategy

This catalog starts with official product and distributor links. Affiliate links can be added later without changing the page design; once a product gets an affiliate URL, the main outbound button can automatically use that tracked commercial link.

© 2026 Black Scarab. AI infrastructure for modern industry.|Based in Miami, FL|info@blackscarab.ai|LinkedIn