CAPTCHA systems are used to differentiate between users and Bots and are used as a security tool for online platforms. There is a need for Devanagari script-based Captcha due to the increased use of content in Devanagari. With the advancement of machine learning algorithms, developing user-friendly and secure Captcha becomes challenging. This paper proposes the Devanagari character recognition in the context of Captcha. It reviews various approaches to Devanagari script recognition, addressing challenges in OCR, segmentation, and feature extraction. Further, it gives the benefits of using the Devanagari script for Captcha. A dataset for Devanagari script-based Captcha is created. The classification algorithm used is KNN because of its simplicity. The algorithm is tested for numerical, vowel, and consonant databases. For numerals and consonants, the accuracy is 98%, whereas for consonants, the accuracy is 61%. The study highlights the use of regional language in Designing Captcha while addressing recognition issues.