diff --git a/cognee/tests/integration/documents/AudioDocument_test.py b/cognee/tests/integration/documents/AudioDocument_test.py
index 9719d90fc..38b547140 100644
--- a/cognee/tests/integration/documents/AudioDocument_test.py
+++ b/cognee/tests/integration/documents/AudioDocument_test.py
@@ -34,7 +34,7 @@ def test_AudioDocument():
     )
     with patch.object(AudioDocument, "create_transcript", return_value=TEST_TEXT):
         for ground_truth, paragraph_data in zip(
-            GROUND_TRUTH, document.read(chunk_size=64, chunker="text_chunker")
+            GROUND_TRUTH, document.read(chunk_size=64, chunker="text_chunker", max_chunk_tokens=512)
         ):
             assert ground_truth["word_count"] == paragraph_data.word_count, (
                 f'{ground_truth["word_count"] = } != {paragraph_data.word_count = }'
diff --git a/cognee/tests/integration/documents/ImageDocument_test.py b/cognee/tests/integration/documents/ImageDocument_test.py
index bd15961ee..faa54fa27 100644
--- a/cognee/tests/integration/documents/ImageDocument_test.py
+++ b/cognee/tests/integration/documents/ImageDocument_test.py
@@ -23,7 +23,7 @@ def test_ImageDocument():
     )
     with patch.object(ImageDocument, "transcribe_image", return_value=TEST_TEXT):
         for ground_truth, paragraph_data in zip(
-            GROUND_TRUTH, document.read(chunk_size=64, chunker="text_chunker")
+            GROUND_TRUTH, document.read(chunk_size=64, chunker="text_chunker", max_chunk_tokens=512)
         ):
             assert ground_truth["word_count"] == paragraph_data.word_count, (
                 f'{ground_truth["word_count"] = } != {paragraph_data.word_count = }'
diff --git a/cognee/tests/integration/documents/PdfDocument_test.py b/cognee/tests/integration/documents/PdfDocument_test.py
index 82d304b6c..e9530fc12 100644
--- a/cognee/tests/integration/documents/PdfDocument_test.py
+++ b/cognee/tests/integration/documents/PdfDocument_test.py
@@ -25,7 +25,7 @@ def test_PdfDocument():
     )
 
     for ground_truth, paragraph_data in zip(
-        GROUND_TRUTH, document.read(chunk_size=1024, chunker="text_chunker")
+        GROUND_TRUTH, document.read(chunk_size=1024, chunker="text_chunker", max_chunk_tokens=2048)
     ):
         assert ground_truth["word_count"] == paragraph_data.word_count, (
             f'{ground_truth["word_count"] = } != {paragraph_data.word_count = }'
diff --git a/cognee/tests/integration/documents/TextDocument_test.py b/cognee/tests/integration/documents/TextDocument_test.py
index 17db39be8..99e28a3ac 100644
--- a/cognee/tests/integration/documents/TextDocument_test.py
+++ b/cognee/tests/integration/documents/TextDocument_test.py
@@ -37,7 +37,8 @@ def test_TextDocument(input_file, chunk_size):
     )
 
     for ground_truth, paragraph_data in zip(
-        GROUND_TRUTH[input_file], document.read(chunk_size=chunk_size, chunker="text_chunker")
+        GROUND_TRUTH[input_file],
+        document.read(chunk_size=chunk_size, chunker="text_chunker", max_chunk_tokens=1024),
     ):
         assert ground_truth["word_count"] == paragraph_data.word_count, (
             f'{ground_truth["word_count"] = } != {paragraph_data.word_count = }'
diff --git a/cognee/tests/integration/documents/UnstructuredDocument_test.py b/cognee/tests/integration/documents/UnstructuredDocument_test.py
index 81e804f07..d76843c0a 100644
--- a/cognee/tests/integration/documents/UnstructuredDocument_test.py
+++ b/cognee/tests/integration/documents/UnstructuredDocument_test.py
@@ -68,7 +68,9 @@ def test_UnstructuredDocument():
     )
 
     # Test PPTX
-    for paragraph_data in pptx_document.read(chunk_size=1024, chunker="text_chunker"):
+    for paragraph_data in pptx_document.read(
+        chunk_size=1024, chunker="text_chunker", max_chunk_tokens=1024
+    ):
         assert 19 == paragraph_data.word_count, f" 19 != {paragraph_data.word_count = }"
         assert 104 == len(paragraph_data.text), f" 104 != {len(paragraph_data.text) = }"
         assert "sentence_cut" == paragraph_data.cut_type, (
@@ -76,7 +78,9 @@ def test_UnstructuredDocument():
         )
 
     # Test DOCX
-    for paragraph_data in docx_document.read(chunk_size=1024, chunker="text_chunker"):
+    for paragraph_data in docx_document.read(
+        chunk_size=1024, chunker="text_chunker", max_chunk_tokens=1024
+    ):
         assert 16 == paragraph_data.word_count, f" 16 != {paragraph_data.word_count = }"
         assert 145 == len(paragraph_data.text), f" 145 != {len(paragraph_data.text) = }"
         assert "sentence_end" == paragraph_data.cut_type, (
@@ -84,7 +88,9 @@ def test_UnstructuredDocument():
         )
 
     # TEST CSV
-    for paragraph_data in csv_document.read(chunk_size=1024, chunker="text_chunker"):
+    for paragraph_data in csv_document.read(
+        chunk_size=1024, chunker="text_chunker", max_chunk_tokens=1024
+    ):
         assert 15 == paragraph_data.word_count, f" 15 != {paragraph_data.word_count = }"
         assert "A A A A A A A A A,A A A A A A,A A" == paragraph_data.text, (
             f"Read text doesn't match expected text: {paragraph_data.text}"
@@ -94,7 +100,9 @@ def test_UnstructuredDocument():
         )
 
     # Test XLSX
-    for paragraph_data in xlsx_document.read(chunk_size=1024, chunker="text_chunker"):
+    for paragraph_data in xlsx_document.read(
+        chunk_size=1024, chunker="text_chunker", max_chunk_tokens=1024
+    ):
         assert 36 == paragraph_data.word_count, f" 36 != {paragraph_data.word_count = }"
         assert 171 == len(paragraph_data.text), f" 171 != {len(paragraph_data.text) = }"
         assert "sentence_cut" == paragraph_data.cut_type, (