diff --git a/tests/utils_test.py b/tests/utils_test.py
index 4ee5d03..58a58e7 100644
--- a/tests/utils_test.py
+++ b/tests/utils_test.py
@@ -116,8 +116,8 @@ def test_batch_text(src, exp, max, min):
             types.TokenDiff(
                 types.TokenDiff.INSERT,
                 '',
-                'example',
-                5,
+                ' example',
+                4,
                 0
             ),
         ],
@@ -128,10 +128,10 @@ def test_batch_text(src, exp, max, min):
         [
             types.TokenDiff(
                 types.TokenDiff.DELETE,
-                'example',
+                ' example',
                 '',
-                5,
-                7
+                4,
+                8
             ),
         ],
     ),
@@ -141,15 +141,15 @@ def test_batch_text(src, exp, max, min):
         [
             types.TokenDiff(
                 types.TokenDiff.DELETE,
-                'example',
+                ' example',
                 '',
-                5,
-                7
+                4,
+                8
             ),
             types.TokenDiff(
                 types.TokenDiff.INSERT,
                 '',
-                'good',
+                'good ',  # XXX: the position of space seems to be a bit inconsistent, before or after
                 18,
                 0
             ),
@@ -167,6 +167,40 @@ def test_batch_text(src, exp, max, min):
         'This is a sentence of 47 characters. ',
         [],
     ),
+    (
+        'This is a sentence.\n'
+        '\n'
+        'This is a new paragraph.\n',
+        'This is a sentence.\n'
+        '\n'
+        'This is the new paragraph.\n',
+        [
+            types.TokenDiff(
+                types.TokenDiff.REPLACE,
+                'a',
+                'the',
+                29,
+                1
+            ),
+        ],
+    ),
+    (
+        'This is a sentence.\n'
+        '\n'
+        'This is a new paragraph.\n',
+        'This is a sentence.\n'
+        '\n'
+        'That this is a new paragraph.\n',
+        [
+            types.TokenDiff(
+                types.TokenDiff.REPLACE,
+                'This',
+                'That this',
+                21,
+                4
+            ),
+        ],
+    ),
 ])
 def test_token_diff(s1, s2, exp):
     res = types.TokenDiff.token_level_diff(s1, s2)
diff --git a/textLSP/analysers/openai/openai.py b/textLSP/analysers/openai/openai.py
index 4f11126..93fbf10 100644
--- a/textLSP/analysers/openai/openai.py
+++ b/textLSP/analysers/openai/openai.py
@@ -85,8 +85,12 @@ def _edit(self, text) -> List[TokenDiff]:
             model=self.config.get(self.CONFIGURATION_MODEL, self.SETTINGS_DEFAULT_MODEL),
             temperature=self.config.get(self.CONFIGURATION_TEMPERATURE, self.SETTINGS_DEFAULT_TEMPERATURE),
         )
+        logger.debug(f"Response: {res}")
+
         if len(res.choices) > 0:
-            return TokenDiff.token_level_diff(text, res.choices[0].message.content.strip())
+            # the API escapes special characters such as newlines
+            res_text = res.choices[0].message.content.strip().encode().decode("unicode_escape")
+            return TokenDiff.token_level_diff(text, res_text)
 
         return []
 
@@ -98,9 +102,11 @@ def _generate(self, text) -> Optional[str]:
             temperature=self.config.get(self.CONFIGURATION_TEMPERATURE, self.SETTINGS_DEFAULT_TEMPERATURE),
             max_tokens=self.config.get(self.CONFIGURATION_MAX_TOKEN, self.SETTINGS_DEFAULT_MAX_TOKEN),
         )
+        logger.debug(f"Response: {res}")
 
         if len(res.choices) > 0:
-            return res.choices[0].message.content.strip()
+            # the API escapes special characters such as newlines
+            return res.choices[0].message.content.strip().encode().decode("unicode_escape")
 
         return None
 
@@ -108,6 +114,9 @@ def _analyse(self, text, doc, offset=0) -> Tuple[List[Diagnostic], List[CodeActi
         diagnostics = list()
         code_actions = list()
 
+        # we don not want trailing whitespace
+        text = text.rstrip()
+
         try:
             edits = self._edit(text)
         except APIError as e:
diff --git a/textLSP/types.py b/textLSP/types.py
index 96da06e..5e5e98a 100644
--- a/textLSP/types.py
+++ b/textLSP/types.py
@@ -317,19 +317,23 @@ class TokenDiff():
     offset: int
     length: int
 
+    @staticmethod
+    def _split(text):
+        return [item for item in re.split("(\s)", text) if item != ""]
+
     @staticmethod
     def token_level_diff(text1, text2) -> List:
-        tokens1 = text1.split()
-        tokens2 = text2.split()
+        tokens1 = TokenDiff._split(text1)
+        tokens2 = TokenDiff._split(text2)
         diff = difflib.SequenceMatcher(None, tokens1, tokens2)
 
         return [
             TokenDiff(
                 type=item[0],
-                old_token=' '.join(tokens1[item[1]:item[2]]),
-                new_token=' '.join(tokens2[item[3]:item[4]]),
-                offset=0 if item[1] == 0 else len(' '.join(tokens1[:item[1]]))+1,
-                length=len(' '.join(tokens1[item[1]:item[2]])),
+                old_token=''.join(tokens1[item[1]:item[2]]),
+                new_token=''.join(tokens2[item[3]:item[4]]),
+                offset=0 if item[1] == 0 else len(''.join(tokens1[:item[1]])),
+                length=len(''.join(tokens1[item[1]:item[2]])),
             )
             for item in diff.get_opcodes()
             if item[0] != 'equal'