diff --git a/bindings/python/py_src/tokenizers/processors/__init__.pyi b/bindings/python/py_src/tokenizers/processors/__init__.pyi
index 5136d02bb..f04806cd6 100644
--- a/bindings/python/py_src/tokenizers/processors/__init__.pyi
+++ b/bindings/python/py_src/tokenizers/processors/__init__.pyi
@@ -98,8 +98,11 @@ class ByteLevel(PostProcessor):
     Args:
         trim_offsets (:obj:`bool`):
             Whether to trim the whitespaces from the produced offsets.
+        add_prefix_space (:obj:`bool`, `optional`, defaults to :obj:`True`):
+            Whether to add a space to the first word if there isn't already one. This
+            lets us treat `hello` exactly like `say hello`.
     """
-    def __init__(self, trim_offsets=True):
+    def __init__(self, trim_offsets=True, add_prefix_space=True):
         pass
 
     def num_special_tokens_to_add(self, is_pair):
diff --git a/bindings/python/src/processors.rs b/bindings/python/src/processors.rs
index 03fa6bdf7..60d4c8ece 100644
--- a/bindings/python/src/processors.rs
+++ b/bindings/python/src/processors.rs
@@ -484,12 +484,16 @@ impl PyRobertaProcessing {
 /// Args:
 ///     trim_offsets (:obj:`bool`):
 ///         Whether to trim the whitespaces from the produced offsets.
+///
+///     add_prefix_space (:obj:`bool`, `optional`, defaults to :obj:`True`):
+///         Whether the add_prefix_space option was enabled during pre-tokenization. This
+///         is relevant because it defines the way the offsets are trimmed out.
 #[pyclass(extends=PyPostProcessor, module = "tokenizers.processors", name = "ByteLevel")]
 pub struct PyByteLevel {}
 #[pymethods]
 impl PyByteLevel {
     #[new]
-    #[pyo3(signature = (add_prefix_space = None, trim_offsets = None, use_regex = None, **_kwargs), text_signature = "(self, trim_offsets=True)")]
+    #[pyo3(signature = (add_prefix_space = None, trim_offsets = None, use_regex = None, **_kwargs), text_signature = "(self, trim_offsets=True, add_prefix_state=True)")]
     fn new(
         add_prefix_space: Option<bool>,
         trim_offsets: Option<bool>,
diff --git a/bindings/python/tests/bindings/test_processors.py b/bindings/python/tests/bindings/test_processors.py
index 3038d8694..c72eb3f6e 100644
--- a/bindings/python/tests/bindings/test_processors.py
+++ b/bindings/python/tests/bindings/test_processors.py
@@ -66,6 +66,7 @@ class TestByteLevelProcessing:
     def test_instantiate(self):
         assert ByteLevel() is not None
         assert ByteLevel(trim_offsets=True) is not None
+        assert ByteLevel(add_prefix_space=True) is not None
         assert isinstance(ByteLevel(), PostProcessor)
         assert isinstance(ByteLevel(), ByteLevel)
         assert isinstance(pickle.loads(pickle.dumps(ByteLevel())), ByteLevel)