Deaf or hard-of-hearing people have been using The American Sign Language (ASL) to communicate with others. Unfortunately, most of the people having normal hearing do not learn such a sign language; therefore, they do not understand persons with such disability. However, the rapid development of science and technology can facilitate people to translate body or part of the body formation more easily. This research is preceded with literature study surveying the need of sensors embedded in a glove. This research employs five flex sensors as well as accelerator and gyroscope to recognize ASL language having similar fingers formation. An Arduino Mega 2560 board as the central controller is employed to read the flex sensors' output and process the information. With 1Sheeld module, the output of the interpreter is presented on a smartphone both in text and voice. The result of this research is a flex glove system capable of translating the ASL from the hand formation that can be seen and be heard. Limitations were found when translating sign for letter N and M as the accuracy reached only 60%; therefore, the total performance of this system to recognize letter A to Z is 96.9%.