We investigate the use of large language models in the screening test of verbal fluency for schizophrenia. Schizophrenia is a mental disorder that affects 1% of the world’s adult population, often characterized by cognitive impairments, including deficits in verbal fluency. Verbal fluency tests are commonly used in clinical practice to evaluate cognitive functions. Our short test includes a standard phonemic verbal fluency task where participants are asked to generate as many words as possible beginning with the letter ’L’. The task is conducted in Slovenian and limited to one minute. We use large language models to assess the words of each participant. In a prompt to the LLM, we describe the task and instruct the LLM to classify each word as either ’correct’, ’repetition’, ’intrusion’, or ’neologism’; the LLM is also instructed to explain its decision for each word. We compare the following models that enable API access: DeepSeek V3, GPT 4-o, LLaMA 3.1, Claude 3.5 and Mistral. An individual prompt includes a list of transcribed words as spoken by our test subjects. We aim to determine the accuracy and reliability of these models, and their potential to be used in clinical practice to assess the cognitive function in schizophrenia.