Monkey tests are usually automated tests with random values that target identified errors in the application. Normally, the errors that emerge in this type of test are due to a failure of treatment in the input or slowness in the treatment of input owing to the stress of the application.
It is common to see the monkey tests used alongside a test technique called Fuzzing. Fuzzing relies on providing invalid, unexpected, and random data as input to computer programs. The program is then monitored, analyzing exceptions such as runtime errors.
In general, the monkey test is useful for identifying out-of-the-box errors.