diff --git a/.github/workflows/tests.yml b/.github/workflows/tests.yml
new file mode 100644
index 0000000..ed1fe26
--- /dev/null
+++ b/.github/workflows/tests.yml
@@ -0,0 +1,47 @@
+name: Tests
+
+on:
+  push:
+    branches:
+      - main
+  pull_request:
+    branches:
+      - main
+
+permissions:
+  contents: read
+
+concurrency:
+  group: >-
+    ${{ github.workflow }}-${{
+      github.event.pull_request.head.repo.full_name || github.repository
+    }}-${{ github.head_ref || github.ref_name }}
+  cancel-in-progress: true
+
+jobs:
+  test:
+    name: Run unit tests
+    runs-on: ubuntu-latest
+    strategy:
+      matrix:
+        python-version: ["3.9", "3.10", "3.11", "3.12"]
+
+    steps:
+      - name: Checkout source code
+        uses: actions/checkout@v4
+
+      - name: Set up Python ${{ matrix.python-version }}
+        uses: actions/setup-python@v5
+        with:
+          python-version: ${{ matrix.python-version }}
+
+      - name: Install dependencies
+        run: |
+          pip install --upgrade pip
+          pip install -r requirements.txt
+
+      - name: Install package
+        run: pip install -e .
+
+      - name: Run tests
+        run: python -m unittest discover -s tests -v
diff --git a/lekcut/attacut.py b/lekcut/attacut.py
index 9d22a81..4937cad 100644
--- a/lekcut/attacut.py
+++ b/lekcut/attacut.py
@@ -203,6 +203,8 @@ def _make_feature(self, txt: str):
         return characters, features
 
     def tokenize(self, text: str) -> List[str]:
+        if not text:
+            return []
         tokens, features = self._make_feature(text)
         logits = self.model.run(None, {"input": features})[0]
         preds = (_sigmoid(logits) > 0.5).astype(int)
@@ -232,6 +234,8 @@ def _make_feature(self, txt: str):
         return characters, features
 
     def tokenize(self, text: str) -> List[str]:
+        if not text:
+            return []
         tokens, features = self._make_feature(text)
         logits = self.model.run(None, {"input": features})[0]
         preds = (_sigmoid(logits) > 0.5).astype(int)
diff --git a/lekcut/deepcut.py b/lekcut/deepcut.py
index 1c09243..8f14195 100644
--- a/lekcut/deepcut.py
+++ b/lekcut/deepcut.py
@@ -138,6 +138,8 @@ def load_model(self, path: str, providers: List[str]=None):
             self.model = ort.InferenceSession(self.path, providers=providers)
     
     def tokenize(self, text: str) -> List[str]:
+        if not text:
+            return []
         self.x_char, self.x_type = create_feature_array(text, n_pad=self.n_pad)
         self.x_char = self.x_char.astype(np.float32)
         self.x_type= self.x_type.astype(np.float32)
diff --git a/tests/__init__.py b/tests/__init__.py
new file mode 100644
index 0000000..e69de29
diff --git a/tests/test_lekcut.py b/tests/test_lekcut.py
new file mode 100644
index 0000000..34d6f97
--- /dev/null
+++ b/tests/test_lekcut.py
@@ -0,0 +1,165 @@
+# -*- coding: utf-8 -*-
+"""Unit tests for the LEKCut Thai word tokenization library."""
+import unittest
+
+from lekcut import word_tokenize
+
+
+class TestWordTokenizeDeepcut(unittest.TestCase):
+    """Tests for the default deepcut model."""
+
+    def test_basic_tokenization(self):
+        result = word_tokenize("ทดสอบการตัดคำ", model="deepcut")
+        self.assertIsInstance(result, list)
+        self.assertTrue(len(result) > 0)
+        self.assertEqual("".join(result), "ทดสอบการตัดคำ")
+
+    def test_known_output(self):
+        result = word_tokenize("ทดสอบการตัดคำ", model="deepcut")
+        self.assertEqual(result, ["ทดสอบ", "การ", "ตัด", "คำ"])
+
+    def test_empty_string(self):
+        result = word_tokenize("", model="deepcut")
+        self.assertEqual(result, [])
+
+    def test_single_word(self):
+        result = word_tokenize("สวัสดี", model="deepcut")
+        self.assertIsInstance(result, list)
+        self.assertEqual("".join(result), "สวัสดี")
+
+    def test_with_spaces(self):
+        result = word_tokenize("สวัสดี ครับ", model="deepcut")
+        self.assertIsInstance(result, list)
+        self.assertEqual("".join(result), "สวัสดี ครับ")
+
+    def test_default_model(self):
+        """word_tokenize defaults to deepcut."""
+        result = word_tokenize("ทดสอบการตัดคำ")
+        self.assertEqual(result, ["ทดสอบ", "การ", "ตัด", "คำ"])
+
+
+class TestWordTokenizeAttacutSC(unittest.TestCase):
+    """Tests for the attacut-sc model."""
+
+    def test_basic_tokenization(self):
+        result = word_tokenize("ทดสอบการตัดคำ", model="attacut-sc")
+        self.assertIsInstance(result, list)
+        self.assertTrue(len(result) > 0)
+        self.assertEqual("".join(result), "ทดสอบการตัดคำ")
+
+    def test_empty_string(self):
+        result = word_tokenize("", model="attacut-sc")
+        self.assertEqual(result, [])
+
+    def test_output_joins_to_input(self):
+        text = "ภาษาไทยสวยงาม"
+        result = word_tokenize(text, model="attacut-sc")
+        self.assertEqual("".join(result), text)
+
+
+class TestWordTokenizeAttacutC(unittest.TestCase):
+    """Tests for the attacut-c model."""
+
+    def test_basic_tokenization(self):
+        result = word_tokenize("ทดสอบการตัดคำ", model="attacut-c")
+        self.assertIsInstance(result, list)
+        self.assertTrue(len(result) > 0)
+        self.assertEqual("".join(result), "ทดสอบการตัดคำ")
+
+    def test_empty_string(self):
+        result = word_tokenize("", model="attacut-c")
+        self.assertEqual(result, [])
+
+    def test_output_joins_to_input(self):
+        text = "ภาษาไทยสวยงาม"
+        result = word_tokenize(text, model="attacut-c")
+        self.assertEqual("".join(result), text)
+
+
+class TestWordTokenizeOskut(unittest.TestCase):
+    """Tests for the oskut model."""
+
+    def test_basic_tokenization(self):
+        result = word_tokenize("ทดสอบการตัดคำ", model="oskut")
+        self.assertIsInstance(result, list)
+        self.assertTrue(len(result) > 0)
+        self.assertEqual("".join(result), "ทดสอบการตัดคำ")
+
+    def test_empty_string(self):
+        result = word_tokenize("", model="oskut")
+        self.assertIsInstance(result, list)
+        self.assertEqual(result, [])
+
+    def test_output_joins_to_input(self):
+        text = "ภาษาไทยสวยงาม"
+        result = word_tokenize(text, model="oskut")
+        self.assertEqual("".join(result), text)
+
+
+class TestWordTokenizeSefrWs1000(unittest.TestCase):
+    """Tests for the sefr-ws1000 model."""
+
+    def test_basic_tokenization(self):
+        result = word_tokenize("ทดสอบการตัดคำ", model="sefr-ws1000")
+        self.assertIsInstance(result, list)
+        self.assertTrue(len(result) > 0)
+        self.assertEqual("".join(result), "ทดสอบการตัดคำ")
+
+    def test_empty_string(self):
+        result = word_tokenize("", model="sefr-ws1000")
+        self.assertEqual(result, [])
+
+    def test_output_joins_to_input(self):
+        text = "ภาษาไทยสวยงาม"
+        result = word_tokenize(text, model="sefr-ws1000")
+        self.assertEqual("".join(result), text)
+
+
+class TestWordTokenizeSefrTnhc(unittest.TestCase):
+    """Tests for the sefr-tnhc model."""
+
+    def test_basic_tokenization(self):
+        result = word_tokenize("ทดสอบการตัดคำ", model="sefr-tnhc")
+        self.assertIsInstance(result, list)
+        self.assertTrue(len(result) > 0)
+        self.assertEqual("".join(result), "ทดสอบการตัดคำ")
+
+    def test_empty_string(self):
+        result = word_tokenize("", model="sefr-tnhc")
+        self.assertEqual(result, [])
+
+    def test_output_joins_to_input(self):
+        text = "ภาษาไทยสวยงาม"
+        result = word_tokenize(text, model="sefr-tnhc")
+        self.assertEqual("".join(result), text)
+
+
+class TestWordTokenizeSefrBest(unittest.TestCase):
+    """Tests for the sefr-best model."""
+
+    def test_basic_tokenization(self):
+        result = word_tokenize("ทดสอบการตัดคำ", model="sefr-best")
+        self.assertIsInstance(result, list)
+        self.assertTrue(len(result) > 0)
+        self.assertEqual("".join(result), "ทดสอบการตัดคำ")
+
+    def test_empty_string(self):
+        result = word_tokenize("", model="sefr-best")
+        self.assertEqual(result, [])
+
+    def test_output_joins_to_input(self):
+        text = "ภาษาไทยสวยงาม"
+        result = word_tokenize(text, model="sefr-best")
+        self.assertEqual("".join(result), text)
+
+
+class TestWordTokenizeErrorHandling(unittest.TestCase):
+    """Tests for error handling in word_tokenize."""
+
+    def test_unsupported_model_raises(self):
+        with self.assertRaises(NotImplementedError):
+            word_tokenize("ทดสอบ", model="unknown-model")
+
+
+if __name__ == "__main__":
+    unittest.main()