# Tiny Machine Learning (Edge AI Glossary)

## metadata
- term: Tiny Machine Learning
- slug: tiny-machine-learning
- letter: T
- route: /resources/glossary/tiny-machine-learning
- canonical: https://klyff.com/resources/glossary/tiny-machine-learning

## Definition & Context
Tiny Machine Learning (TinyML) is the practice of deploying machine learning models directly on edge devices, enabling real-time intelligence without cloud dependency. Also called...

## Details
## Tiny Machine Learning

#### What is Tiny Machine Learning?

Tiny Machine Learning (TinyML) is the practice of deploying machine learning models directly on edge devices, enabling real-time intelligence without cloud dependency. Also called “micro AI,” it brings smart capabilities to small, low-power hardware at the network’s edge.

#### Why Is It Used?

TinyML is used to reduce latency, save bandwidth, enhance privacy, and enable autonomous decision-making in resource-constrained environments like IoT sensors, wearables, and industrial devices.

#### How Is It Used?

 
- Running predictive models locally on microcontrollers or sensors
- Performing real-time anomaly detection, audio/visual recognition, and environmental monitoring
- Power-efficient on-device analytics without sending data to the cloud 

#### Types of Tiny Machine Learning

 
- **On-Device TinyML:** Fully operates on microcontrollers or edge hardware
- **Hybrid TinyML:** Combines on-device inference with cloud-based model updates
- **Streaming TinyML:** Processes continuous sensor data for real-time predictions 

#### Benefits of Tiny Machine Learning

 
- Ultra-low latency responses
- Reduced cloud costs and network dependency
- Improved data privacy and security
- Energy-efficient AI suitable for battery-powered devices