diff --git a/README.md b/README.md
index e420a7b..a0c530b 100644
--- a/README.md
+++ b/README.md
@@ -11,7 +11,8 @@ Simple streaming JSON parser and encoder.
 When [reading](#reading) JSON data, `json-stream` can decode JSON data in 
 a streaming manner, providing a pythonic dict/list-like interface, or a
 [visitor-based interface](#visitor). It can stream from files, [URLs](#urls) 
-or [iterators](#iterators). It can process [multiple JSON documents](#multiple) in a single stream.
+or [iterators](#iterators). It can process [multiple JSON documents](#multiple) 
+in a single stream, and can read JSON [mixed with other non-JSON data](#reading-mixed-data).
 
 When [writing](#writing) JSON data, `json-stream` can stream JSON objects 
 as you generate them.
@@ -495,13 +496,81 @@ significant parsing speedup compared to pure python implementation.
 `json-stream` will fallback to its pure python tokenizer implementation
 if `json-stream-rs-tokenizer` is not available.
 
+#### <a id="reading-mixed-data"></a> Reading mixed data
+
+When using the Rust tokenizer, you can also use `json-stream` to parse mixed
+data, for example a file with a JSON followed by binary data. 
+ 
+To do this, you should pass `correct_cursor=True` to `load()`. The ensures the
+rust tokenizer keeps track of the exact stream position it has read up to. This
+comes with a **significant performance cost** for un-seekable streams.
+
+After reading the JSON data, call `read_all()` on the top-level object returned
+by `load()` to ensure you have read up to the end of the JSON data, and then call 
+`.tokenizer.park_cursor()` to "park" the underlying file cursor at the correct
+position.
+
+```python
+import json_stream
+
+with open('test.bin', 'rb') as f:
+    # read JSON header
+    header = json_stream.load(f, correct_cursor=True)
+    # ... process JSON header ...
+    header.read_all()
+
+    # ensure the tokenizer has "parked" the file 
+    # cursor at the end of the JSON data
+    header.tokenizer.park_cursor()
+
+    # now we can read binary data from the same file
+    binary_start = f.tell()
+    data = f.read()
+
+#### <a id="mixed-scenarios"></a> Other mixed data scenarios
+
+`json-stream` can also handle streams that start with binary data, or have binary
+data between multiple JSON documents.
+
+##### Binary then JSON
+
+You can simply read the binary data from the file before calling `load()`.
+
+```python
+with open('test.bin', 'rb') as f:
+    binary_data = f.read(1024)
+    data = json_stream.load(f)
+    # ... process JSON ...
+```
+
+##### JSON then binary then JSON
+
+You must use `correct_cursor=True` for any JSON document that is followed by 
+binary data.
+
+```python
+with open('test.bin', 'rb') as f:
+    # 1. Read first JSON
+    data1 = json_stream.load(f, correct_cursor=True)
+    # ... process data1 ...
+    data1.read_all()
+    data1.tokenizer.park_cursor()
+
+    # 2. Read binary data
+    binary_data = f.read(1024)
+
+    # 3. Read second JSON
+    data2 = json_stream.load(f)
+    # ... process data2 ...
+```
+
 ### Custom tokenizer
 
 You can supply an alternative JSON tokenizer implementation. Simply pass 
 a tokenizer to the `load()` or `visit()` methods.
 
 ```python
-json_stream.load(f, tokenizer=some_tokenizer)
+json_stream.load(f, tokenizer=some_tokenizer, **tokenizer_kwargs)
 ```
 
 The requests methods also accept a customer tokenizer parameter.
diff --git a/src/json_stream/base.py b/src/json_stream/base.py
index c5e3abd..6f113c2 100644
--- a/src/json_stream/base.py
+++ b/src/json_stream/base.py
@@ -36,6 +36,10 @@ def __init__(self, token_stream):
         self._stream = token_stream
         self._child: Optional[StreamingJSONBase] = None
 
+    @property
+    def tokenizer(self):
+        return self._stream
+
     def _clear_child(self):
         if self._child is not None:
             self._child.read_all()
diff --git a/src/json_stream/loader.py b/src/json_stream/loader.py
index a1493bf..cda0cbe 100644
--- a/src/json_stream/loader.py
+++ b/src/json_stream/loader.py
@@ -3,13 +3,13 @@
 from json_stream.select_tokenizer import default_tokenizer
 
 
-def load(fp_or_iterable, persistent=False, tokenizer=default_tokenizer):
-    return next(load_many(fp_or_iterable, persistent, tokenizer))
+def load(fp_or_iterable, persistent=False, tokenizer=default_tokenizer, **tokenizer_kwargs):
+    return next(load_many(fp_or_iterable, persistent, tokenizer, **tokenizer_kwargs))
 
 
-def load_many(fp_or_iterable, persistent=False, tokenizer=default_tokenizer):
+def load_many(fp_or_iterable, persistent=False, tokenizer=default_tokenizer, **tokenizer_kwargs):
     fp = ensure_file(fp_or_iterable)
-    token_stream = tokenizer(fp)
+    token_stream = tokenizer(fp, **tokenizer_kwargs)
     for token_type, token in token_stream:
         if token_type == TokenType.OPERATOR:
             data = StreamingJSONBase.factory(token, token_stream, persistent)
diff --git a/src/json_stream/tests/test_binary_resumption.py b/src/json_stream/tests/test_binary_resumption.py
new file mode 100644
index 0000000..22f91d7
--- /dev/null
+++ b/src/json_stream/tests/test_binary_resumption.py
@@ -0,0 +1,97 @@
+import io
+import json
+from unittest import TestCase, skipUnless
+import json_stream
+
+try:
+    import json_stream_rs_tokenizer
+    HAS_RS_TOKENIZER = hasattr(json_stream_rs_tokenizer, 'RustTokenizer')
+except ImportError:
+    HAS_RS_TOKENIZER = False
+
+class TestBinaryResumption(TestCase):
+    @skipUnless(HAS_RS_TOKENIZER, 'Rust tokenizer not available')
+    def test_json_then_binary(self):
+        json_header = json.dumps({"header": "info"})
+        binary_data = b'\x00\x01\x02\x03'
+        test_data = json_header.encode('utf-8') + binary_data
+        test_file = io.BytesIO(test_data)
+        
+        # Load with correct_cursor=True
+        header = json_stream.load(test_file, correct_cursor=True)
+        
+        # Consume all data from header
+        header.read_all()
+        
+        # Signal that we are done with JSON and want to resume binary read
+        header.tokenizer.park_cursor()
+        
+        # Verify file cursor position
+        self.assertEqual(test_file.tell(), len(json_header))
+        
+        # Verify binary data
+        remaining = test_file.read()
+        self.assertEqual(remaining, binary_data)
+
+    @skipUnless(HAS_RS_TOKENIZER, 'Rust tokenizer not available')
+    def test_binary_then_json(self):
+        binary_data = b'binary_start'
+        json_data = b'{"a": 1}'
+        test_data = binary_data + json_data
+        test_file = io.BytesIO(test_data)
+        
+        # Read binary
+        read_binary = test_file.read(len(binary_data))
+        self.assertEqual(read_binary, binary_data)
+        
+        # Load JSON
+        data = json_stream.load(test_file)
+        self.assertEqual(dict(data.items()), {"a": 1})
+
+    @skipUnless(HAS_RS_TOKENIZER, 'Rust tokenizer not available')
+    def test_json_then_binary_then_json(self):
+        json_1 = b'{"first": true}'
+        binary_middle = b'middle_binary'
+        json_2 = b'{"second": false}'
+        test_data = json_1 + binary_middle + json_2
+        test_file = io.BytesIO(test_data)
+        
+        # Load first JSON
+        data1 = json_stream.load(test_file, correct_cursor=True)
+        self.assertEqual(dict(data1.items()), {"first": True})
+        data1.read_all()
+        data1.tokenizer.park_cursor()
+        self.assertEqual(test_file.tell(), len(json_1))
+        
+        # Read middle binary
+        read_middle = test_file.read(len(binary_middle))
+        self.assertEqual(read_middle, binary_middle)
+        
+        # Load second JSON
+        data2 = json_stream.load(test_file)
+        self.assertEqual(dict(data2.items()), {"second": False})
+
+    @skipUnless(HAS_RS_TOKENIZER, 'Rust tokenizer not available')
+    def test_load_many_then_binary(self):
+        json_1 = '{"a": 1}'
+        json_2 = '{"b": 2}'
+        binary_data = b'binary'
+        test_data = json_1.encode('utf-8') + json_2.encode('utf-8') + binary_data
+
+        test_file = io.BytesIO(test_data)
+
+        loader = json_stream.load_many(test_file, correct_cursor=True)
+
+        # Read first JSON
+        doc1 = next(loader)
+        doc1.read_all()
+
+        # Read second JSON
+        doc2 = next(loader)
+        doc2.read_all()
+
+        # Now park cursor
+        doc2.tokenizer.park_cursor()
+
+        self.assertEqual(test_file.tell(), len(json_1) + len(json_2))
+        self.assertEqual(test_file.read(), binary_data)
diff --git a/src/json_stream/visitor.py b/src/json_stream/visitor.py
index b321670..392919a 100644
--- a/src/json_stream/visitor.py
+++ b/src/json_stream/visitor.py
@@ -20,9 +20,9 @@ def _visit(obj, visitor, path):
         visitor(obj, path)
 
 
-def visit_many(fp_or_iterator, visitor, tokenizer=default_tokenizer):
+def visit_many(fp_or_iterator, visitor, tokenizer=default_tokenizer, **tokenizer_kwargs):
     fp = ensure_file(fp_or_iterator)
-    token_stream = tokenizer(fp)
+    token_stream = tokenizer(fp, **tokenizer_kwargs)
     for token_type, token in token_stream:
         if token_type == TokenType.OPERATOR:
             obj = StreamingJSONBase.factory(token, token_stream, persistent=False)
@@ -33,5 +33,5 @@ def visit_many(fp_or_iterator, visitor, tokenizer=default_tokenizer):
         yield
 
 
-def visit(fp_or_iterator, visitor, tokenizer=default_tokenizer):
-    next(visit_many(fp_or_iterator, visitor, tokenizer))
+def visit(fp_or_iterator, visitor, tokenizer=default_tokenizer, **tokenizer_kwargs):
+    next(visit_many(fp_or_iterator, visitor, tokenizer, **tokenizer_kwargs))