Phishing attacks are a prevalent security threat, and detecting and preventing such attacks is crucial to safeguarding sensitive information. By performing aphishing attack the attackercangetholdofthevictim’spersonaldetailsincludinglogincredentials,andcreditcarddetails,andperformso mefraudulentactivities.To addressthisissue,ourproposedmethodmakes use of machine learning techniques and uses so me classification algorithms, such as K-nearest neighbor, decision trees, Random Forest and Ada Boost to identify phishing urls. For this we use a data set that consists of 38,625data of which 16,252dataarelegitimateandaretakenfromalexa.comand22,373dataarephishingtakenfrom phishtank.com. The data pre-processing is performed on the data by applying techniques such as under sampling and over-sampling, and as a partoffeatureextraction12featuresareselected and the model is trained on these data, then the model is tested using the test data. Finally, we evaluate the performance of each algorithm using performance metrics such a saccuracy, precision, f1score,and recall. After evaluating the algorithms, we save the best-performing model in a pickle file. Our results indicate that the Random Forest frame work, we developed a web application, where the user can check the legiti macy U of the URL. Once the user enters the URL in the search bar provided, then our model will predictwhethertheURLislegitimateoraphishingattempt,andifitisaphishing URL,awarning message will be displayed to the user. This approach will help prevent users from falling victim to phishing attacks and safe guard their sensitive information.