diff --git a/README.md b/README.md
index 5e24b2d..db86f9b 100644
--- a/README.md
+++ b/README.md
@@ -464,6 +464,34 @@ Click [here](https://github.com/beehive-lab/GPULlama3.java/tree/main/docs/TORNAD
 
 Click [here](https://github.com/beehive-lab/GPULlama3.java/tree/main/docs/GPULlama3_ROADMAP.md) to see the roadmap of the project.
 
+
+## Run All Tests
+
+You can run all unit tests using the following Maven command:
+
+```bash
+  mvn test
+
+Sample Output
+-------------------------------------------------------
+ T E S T S
+-------------------------------------------------------
+Running com.example.tokenizer.impl.MistralTokenizerTest
+Running com.example.tokenizer.impl.TokenizerInterfaceTest
+
+Tests run: 12, Failures: 0, Errors: 0, Skipped: 0
+
+To run tests inside an IDE (e.g., IntelliJ), right-click on the test classes and choose Run.
+
+## Test Coverage
+Here are the tokenizer unit tests included:
+
+| **Test Class**           | **Description**                                                                                                                             |
+|--------------------------|---------------------------------------------------------------------------------------------------------------------------------------------|
+| `MistralTokenizerTest`   | Verifies Mistral tokenizer functionality including byte fallback (`<0xXX>`), special token handling, encoding and decoding logic           |
+| `TokenizerInterfaceTest` | Unit tests for utility methods like `replaceControlCharacters`, ensuring printable and safe token rendering                                |
+
+```
 -----------
 
 ## Acknowledgments
diff --git a/pom.xml b/pom.xml
index 216dda9..0908f89 100644
--- a/pom.xml
+++ b/pom.xml
@@ -32,6 +32,12 @@
             <artifactId>tornado-runtime</artifactId>
             <version>1.1.1-dev</version>
         </dependency>
+        <dependency>
+            <groupId>org.junit.jupiter</groupId>
+            <artifactId>junit-jupiter</artifactId>
+            <version>5.10.2</version>
+            <scope>test</scope>
+        </dependency>
     </dependencies>
 
     <build>
@@ -68,6 +74,11 @@
                     </execution>
                 </executions>
             </plugin>
+            <plugin>
+                <groupId>org.apache.maven.plugins</groupId>
+                <artifactId>maven-surefire-plugin</artifactId>
+                <version>3.2.5</version>
+            </plugin>
         </plugins>
     </build>
 </project>
diff --git a/src/test/java/com/example/tokenizer/impl/MistralTokenizerTest.java b/src/test/java/com/example/tokenizer/impl/MistralTokenizerTest.java
new file mode 100644
index 0000000..81e365e
--- /dev/null
+++ b/src/test/java/com/example/tokenizer/impl/MistralTokenizerTest.java
@@ -0,0 +1,80 @@
+package com.example.tokenizer.impl;
+
+import com.example.tokenizer.vocabulary.Vocabulary;
+import org.junit.jupiter.api.BeforeEach;
+import org.junit.jupiter.api.Test;
+
+import java.util.*;
+
+import static org.junit.jupiter.api.Assertions.*;
+
+class MistralTokenizerTest {
+
+    private Vocabulary vocabulary;
+    private MistralTokenizer tokenizer;
+
+    @BeforeEach
+    void setup() {
+        List<String> baseTokens = List.of("▁h", "e", "l", "o", "▁", "▁hello");
+        List<String> byteFallbackTokens = new ArrayList<>();
+
+        for (int i = 0; i < 256; i++) {
+            byteFallbackTokens.add(String.format("<0x%02X>", i));
+        }
+
+        List<String> allTokens = new ArrayList<>();
+        allTokens.addAll(baseTokens);
+        allTokens.addAll(byteFallbackTokens);
+
+        String[] tokens = allTokens.toArray(new String[0]);
+        float[] scores = new float[tokens.length];
+        Arrays.fill(scores, 0.0f); // dummy scores
+
+        int[] tokenTypes = new int[tokens.length];
+        Arrays.fill(tokenTypes, 1); // mark all normal
+        tokenTypes[baseTokens.size()] = 0; // mark <0x00> as special
+
+        Map<String, Object> metadata = new HashMap<>();
+        metadata.put("tokenizer.ggml.token_type", tokenTypes);
+
+        vocabulary = new Vocabulary(tokens, scores);
+        tokenizer = new MistralTokenizer(metadata, vocabulary);
+    }
+
+    @Test
+    void testEncodeSimpleText() {
+        List<Integer> tokens = tokenizer.encodeAsList("hello");
+        assertNotNull(tokens);
+        assertFalse(tokens.isEmpty());
+    }
+
+    @Test
+    void testRegexPatternReturnsNull() {
+        assertNull(tokenizer.regexPattern());
+    }
+
+    @Test
+    void testSpecialTokenDetection() {
+        assertTrue(tokenizer.isSpecialToken(6));
+        assertFalse(tokenizer.isSpecialToken(0));
+    }
+
+    @Test
+    void testShouldDisplayToken() {
+        assertTrue(tokenizer.shouldDisplayToken(0));
+        assertFalse(tokenizer.shouldDisplayToken(6));
+    }
+
+    @Test
+    void testDecodeSpecialByteFallbackToken() {
+        List<Integer> tokens = List.of(6); // token <0x00>
+        String result = tokenizer.decode(tokens);
+        assertEquals("\u0000", result); // ASCII for <0x00>
+    }
+
+    @Test
+    void testEncodeEmptyInput() {
+        List<Integer> tokens = tokenizer.encodeAsList("");
+        assertTrue(tokens.isEmpty(), "Should return empty token list for empty input");
+    }
+}
\ No newline at end of file
diff --git a/src/test/java/com/example/tokenizer/impl/TokenizerTest.java b/src/test/java/com/example/tokenizer/impl/TokenizerTest.java
new file mode 100644
index 0000000..dce2e46
--- /dev/null
+++ b/src/test/java/com/example/tokenizer/impl/TokenizerTest.java
@@ -0,0 +1,56 @@
+package com.example.tokenizer.impl;
+
+import org.junit.jupiter.api.Test;
+
+import static org.junit.jupiter.api.Assertions.*;
+
+class TokenizerInterfaceTest {
+
+    @Test
+    void testReplaceControlCharactersWithCodePoints() {
+        int[] input = {'H', 'e', '\n', 0x07, 'l', 'o'}; // 0x07 = BEL (control character)
+        String result = Tokenizer.replaceControlCharacters(input);
+
+        assertEquals("He\n\\u0007lo", result); // \n allowed, BEL escaped
+    }
+
+    @Test
+    void testReplaceControlCharactersWithString() {
+        String input = "He\n\u0007lo"; // \u0007 is a bell character (non-printable control char)
+        String result = Tokenizer.replaceControlCharacters(input);
+
+        assertEquals("He\n\\u0007lo", result);
+    }
+
+    @Test
+    void testReplaceControlCharactersWithOnlyPrintableChars() {
+        String input = "Hello, World!";
+        String result = Tokenizer.replaceControlCharacters(input);
+
+        assertEquals(input, result);
+    }
+
+    @Test
+    void testReplaceControlCharactersWithMultipleControlChars() {
+        String input = "\u0001\u0002A\nB\u0003"; // \u0001, \u0002, \u0003 are control chars
+        String result = Tokenizer.replaceControlCharacters(input);
+
+        assertEquals("\\u0001\\u0002A\nB\\u0003", result);
+    }
+
+    @Test
+    void testReplaceControlCharactersEmptyInput() {
+        String input = "";
+        String result = Tokenizer.replaceControlCharacters(input);
+
+        assertEquals("", result);
+    }
+
+    @Test
+    void testReplaceControlCharactersNullSafe() {
+        // Add this test if you plan to make it null-safe.
+        assertThrows(NullPointerException.class, () -> {
+            Tokenizer.replaceControlCharacters((String) null);
+        });
+    }
+}
\ No newline at end of file