diff --git a/README.md b/README.md
index ab4d08b..015dc56 100644
--- a/README.md
+++ b/README.md
@@ -396,7 +396,18 @@ Portuguese and `es_419` Latin American Spanish; plain `pt` and `es` stay
 unqualified, as does `ar`). The optional `ducking` boolean (default off, free) ducks the
 background music/effects bed under the dubbed voice while it speaks; when off
 the bed is kept at a constant level. (Every endpoint's `ducking` defaults off,
-so this one is no exception.) Source videos may be
+so this one is no exception.)
+
+The optional `lipsync` boolean is the one parameter here that defaults **on**:
+the speaker's mouth is re-rendered to match the dubbed speech. Pass
+`lipsync=False` to leave the picture completely untouched instead — the video
+comes back at its original resolution and frame rate rather than re-rendered,
+and only the audio is replaced, so the mouths keep moving to the original
+language. Reach for it on footage with no on-camera speaker, or when
+preserving the exact original picture matters more than matching lip movement.
+The background bed is rebuilt either way, so `ducking` behaves the same.
+
+Source videos may be
 at most 300 seconds long, and billing is per language: a 3-language call
 costs three times as much as one. Dubbing has no free trial allowance — see
 [Free trial](#free-trial).
diff --git a/sonilo-cli/README.md b/sonilo-cli/README.md
index d0bef08..a22e90f 100644
--- a/sonilo-cli/README.md
+++ b/sonilo-cli/README.md
@@ -277,6 +277,11 @@ command that produces no media file — nothing is generated:
   tr, vi, id` (`pt_br` is Brazilian Portuguese and `es_419` Latin American
   Spanish; plain `pt` and `es` stay unqualified, as does `ar`).
 - Source videos may be at most 300 seconds long.
+- `--no-lipsync` leaves the picture completely untouched. By default the speaker's mouth is
+  re-rendered to match the dubbed speech; with this flag the video comes back at its original
+  resolution and frame rate and only the audio is replaced, so the mouths keep moving to the
+  original language. Use it for footage with no on-camera speaker, or when preserving the exact
+  original picture matters more than matching lip movement.
 - `--output` is a filename template, not a single destination: a dubbing task returns one video
   per language, so `--output clip.mp4` writes `clip.es.mp4`, `clip.fr.mp4`, etc.
 - Billing is per language, and dubbing has **no free trial runs** — see [Free trial](#free-trial)
diff --git a/sonilo-cli/src/sonilo_cli/__main__.py b/sonilo-cli/src/sonilo_cli/__main__.py
index e42ff6b..6bb8d7d 100644
--- a/sonilo-cli/src/sonilo_cli/__main__.py
+++ b/sonilo-cli/src/sonilo_cli/__main__.py
@@ -635,6 +635,9 @@ def cmd_dubbing(client: Sonilo, args: argparse.Namespace) -> None:
         video=args.video,
         video_url=args.video_url,
         languages=languages,
+        # Only sent when the flag is present, so the server keeps owning the
+        # default (lip sync on).
+        lipsync=False if args.no_lipsync else None,
         timeout=args.timeout,
     )
     if not result.outputs:
@@ -1010,6 +1013,16 @@ def build_parser() -> argparse.ArgumentParser:
              "es_419 Latin American Spanish; plain pt and es stay "
              "unqualified, as does ar.",
     )
+    p_dub.add_argument(
+        "--no-lipsync", dest="no_lipsync", action="store_true",
+        help="Leave the picture completely untouched. By default the speaker's "
+             "mouth is re-rendered to match the dubbed speech; with this the "
+             "video comes back at its original resolution and frame rate and "
+             "only the audio is replaced, so the mouths keep moving to the "
+             "original language. Use it for footage with no on-camera speaker, "
+             "or when preserving the exact original picture matters more than "
+             "matching lip movement.",
+    )
     p_dub.add_argument(
         "--output", default=None,
         help="Filename template; one file is written per language with the code "
diff --git a/src/sonilo/_requests.py b/src/sonilo/_requests.py
index d36e58a..fdfcbaa 100644
--- a/src/sonilo/_requests.py
+++ b/src/sonilo/_requests.py
@@ -174,6 +174,7 @@ def build_dubbing_parts(
     video_url: Optional[str],
     languages: Optional[List[str]],
     ducking: Optional[bool] = None,
+    lipsync: Optional[bool] = None,
 ) -> Tuple[Dict[str, str], Optional[Dict[str, tuple]], bool]:
     """Build the multipart parts for POST /v1/dubbing.
 
@@ -206,6 +207,11 @@ def build_dubbing_parts(
     # when unset so the server default applies.
     if ducking is not None:
         data["ducking"] = "true" if ducking else "false"
+    # Default-ON server-side, unlike ducking — this is the one parameter here
+    # whose useful direction is turning it off. Omitted when unset all the
+    # same, so the server keeps owning the default.
+    if lipsync is not None:
+        data["lipsync"] = "true" if lipsync else "false"
 
     # Now open files (only after data is fully assembled)
     files: Optional[Dict[str, tuple]] = None
diff --git a/src/sonilo/resources/dubbing.py b/src/sonilo/resources/dubbing.py
index d5b5adf..3ca409e 100644
--- a/src/sonilo/resources/dubbing.py
+++ b/src/sonilo/resources/dubbing.py
@@ -25,7 +25,16 @@ class Dubbing:
     `ducking` (default off, free) ducks the background music/effects bed
     under the dubbed voice while it speaks; when off the bed is kept at a
     constant level. Every endpoint's `ducking` defaults off, so this one is
-    no exception."""
+    no exception.
+
+    `lipsync` is the one parameter here that defaults **on**: the speaker's
+    mouth is re-rendered to match the dubbed speech. Pass `lipsync=False` to
+    leave the picture completely untouched instead — the video comes back at
+    its original resolution and frame rate rather than re-rendered, and only
+    the audio is replaced, so the mouths keep moving to the original language.
+    Reach for it on footage with no on-camera speaker, or when preserving the
+    exact original picture matters more than matching lip movement. The
+    background bed is rebuilt either way, so `ducking` behaves the same."""
 
     def __init__(self, client: "Sonilo") -> None:
         self._client = client
@@ -37,8 +46,11 @@ def submit(
         video_url: Optional[str] = None,
         languages: Optional[List[str]] = None,
         ducking: Optional[bool] = None,
+        lipsync: Optional[bool] = None,
     ) -> SfxTask:
-        data, files, opened = build_dubbing_parts(video, video_url, languages, ducking)
+        data, files, opened = build_dubbing_parts(
+            video, video_url, languages, ducking, lipsync
+        )
         close_after = files["video"][1] if files is not None and opened else None
         return parse_sfx_task(
             self._client._post_json(PATH, data=data, files=files, close_after=close_after)
@@ -51,11 +63,13 @@ def generate(
         video_url: Optional[str] = None,
         languages: Optional[List[str]] = None,
         ducking: Optional[bool] = None,
+        lipsync: Optional[bool] = None,
         poll_interval: float = DEFAULT_POLL_INTERVAL,
         timeout: float = DEFAULT_WAIT_TIMEOUT,
     ) -> DubbingResult:
         task = self.submit(
-            video=video, video_url=video_url, languages=languages, ducking=ducking
+            video=video, video_url=video_url, languages=languages,
+            ducking=ducking, lipsync=lipsync,
         )
         return self._client.tasks.wait(
             task.task_id,
@@ -76,8 +90,11 @@ async def submit(
         video_url: Optional[str] = None,
         languages: Optional[List[str]] = None,
         ducking: Optional[bool] = None,
+        lipsync: Optional[bool] = None,
     ) -> SfxTask:
-        data, files, opened = build_dubbing_parts(video, video_url, languages, ducking)
+        data, files, opened = build_dubbing_parts(
+            video, video_url, languages, ducking, lipsync
+        )
         close_after = files["video"][1] if files is not None and opened else None
         return parse_sfx_task(
             await self._client._post_json(
@@ -92,11 +109,13 @@ async def generate(
         video_url: Optional[str] = None,
         languages: Optional[List[str]] = None,
         ducking: Optional[bool] = None,
+        lipsync: Optional[bool] = None,
         poll_interval: float = DEFAULT_POLL_INTERVAL,
         timeout: float = DEFAULT_WAIT_TIMEOUT,
     ) -> DubbingResult:
         task = await self.submit(
-            video=video, video_url=video_url, languages=languages, ducking=ducking
+            video=video, video_url=video_url, languages=languages,
+            ducking=ducking, lipsync=lipsync,
         )
         return await self._client.tasks.wait(
             task.task_id,
diff --git a/tests/test_dubbing.py b/tests/test_dubbing.py
index ddd5706..41b843b 100644
--- a/tests/test_dubbing.py
+++ b/tests/test_dubbing.py
@@ -119,6 +119,24 @@ def test_submit_sends_ducking_only_when_set():
     assert "ducking" not in bodies[2]
 
 
+@respx.mock
+def test_submit_sends_lipsync_only_when_set():
+    """The mirror of ducking, with the default the other way up: absent must
+    mean lip sync ON, which is what every dubbing call did before the
+    parameter existed."""
+    route = respx.post("https://api.sonilo.com/v1/dubbing").mock(
+        return_value=httpx.Response(202, json=ACK)
+    )
+    with Sonilo(api_key="sk-test") as client:
+        client.dubbing.submit(video_url="https://x/v.mp4", lipsync=False)
+        client.dubbing.submit(video_url="https://x/v.mp4", lipsync=True)
+        client.dubbing.submit(video_url="https://x/v.mp4")
+    bodies = [unquote_plus(c.request.content.decode()) for c in route.calls]
+    assert "lipsync=false" in bodies[0]
+    assert "lipsync=true" in bodies[1]
+    assert "lipsync" not in bodies[2]
+
+
 @respx.mock
 def test_generate_polls_to_a_dubbing_result():
     respx.post("https://api.sonilo.com/v1/dubbing").mock(