diff --git a/tests/describe.PY b/tests/describe.PY
new file mode 100755
index 0000000..4e1d844
--- /dev/null
+++ b/tests/describe.PY
@@ -0,0 +1,69 @@
+#! /Library/Frameworks/Python.framework/Versions/3.12/bin/python3
+# describe.py
+# To fulfill the test of asking the LLMs to describe
+
+# Import modules. 
+import ollama
+import json
+
+# Add source files. 
+IMPORTED = {'Strings': "data/datasets/strings.JSON", 'Prompts': "tests/config/prompts.json", 'Models': 'tests/config/models.JSON'}
+ 
+# Set up the main variables. 
+RESULTS = {};
+
+# Read the files. 
+for NAME in list(IMPORTED.keys()): 
+    try: 
+        DATA = json.load(open(IMPORTED[NAME]))
+    except: 
+        DATA = open(IMPORTED[NAME]).read()
+
+    IMPORTED[NAME] = DATA;
+
+# Download the models. 
+def download_models(): 
+    for MODEL_NAME in IMPORTED["Models"].keys(): 
+        MODEL_ID = IMPORTED["Models"][MODEL_NAME];
+        ollama.pull(MODEL_ID);
+
+def format_prompt(): 
+    PROMPT = f"{IMPORTED['Prompts']["sample"]}\n";
+
+    for GENERATION_TYPE in IMPORTED['Strings']['training'].keys(): 
+        for TEXT_NUMBER in range(len(IMPORTED['Strings']['training'][GENERATION_TYPE])): 
+            PROMPT = f"{PROMPT}\n\n{GENERATION_TYPE}-written #{str(TEXT_NUMBER + 1)}: \n“{'\n\n\t'.join(IMPORTED['Strings']['training'][GENERATION_TYPE][TEXT_NUMBER].strip().split("\n\n"))}”";
+
+    PROMPT = f"{PROMPT}\n\n{IMPORTED['Prompts']["description"]}";
+
+    return (PROMPT);
+
+# Execute the response. 
+def ask_AI(PROMPT): 
+    if len(list(IMPORTED['Models'].keys())): 
+
+        for MODEL_NAME in list(IMPORTED['Models'].keys()): 
+            MODEL_ID = IMPORTED['Models'][MODEL_NAME];
+            RESULTS[MODEL_NAME] = (ollama.generate(model=MODEL_ID, prompt=PROMPT))['response'].strip();
+
+    return (RESULTS);
+
+# Display the outputs. 
+# Save all of the strings. 
+# Parameters: 
+#   filename: The file name
+def save_data(**parameters): 
+    OUTPUT = parameters['dictionary'];
+
+    if (parameters['filename'].strip()): 
+        with open(parameters['filename'], 'w') as file:
+            json.dump(OUTPUT, file);
+
+# Execute the script. 
+download_models();
+PROMPT = format_prompt();
+print(f"Using prompt: \n\t|\t{'\n\t|\t'.join(PROMPT.split("\n"))}");
+RESULTS = ask_AI(PROMPT);
+save_data(dictionary=RESULTS, filename='tests/outputs/descriptions.JSON')
+
+