Gesture recognition using TinyML and Wio Terminal is a project where machine learning models have been used to recognise gestures in real-time. Use of machine learning and microcontrollers in this project has opened up opportunities for various applications in fields such as human-computer interaction, robotics, VR & AR gaming experience, Web3, spatial computing technology, and many others. A Convolutional Neural Network (CNN) model that has been trained to identify hand gestures from a dataset of real-world gestures serves as the foundation for this research. The dataset is processed to extract relevant features, and the model is trained using Codecraft, a popular tiny machine learning application. The Wio Terminal, a compact and versatile device, is used to capture live input from the user's hand gestures using its built-in accelerometer and gyroscope sensors. The detected gestures are classified in realtime using the deployed machine learning model and the corresponding actions are displayed on the Wio Terminal's LCD screen. The system is a workable option for numerous real-time gesture recognition applications due to its effective resource management and small footprint in wastage of resources. The project's outcomes show that real-time gesture detection applications employing TinyML are feasible, opening up a larger spectrum of possibilities. Compact, affordable, and low-power gesture recognition systems that may be employed in a variety of applications can now be built using machine learning and microcontrollers.The demonstration also illustrates the potential of machine learning models in embedded systems, which may pave the way for the future creation of more sophisticated systems and applications in nano technology using ML.