feat: fast mode on lucy-vton-3.6; lucy-vton-latest resolves to 3.6

README.md

@@ -120,9 +120,10 @@ code use `asyncio.run(verify_client_token(token))`.
 Realtime sessions accept an optional `speed` on `RealtimeConnectOptions`, alongside `resolution`.
 Fast mode (`speed="fast"`) serves the session from a higher-compute tier for lower latency and
 higher throughput; output quality is unchanged. It is currently available for `lucy-2.5` /
-`lucy-latest` and `lucy-vton-3.5` / `lucy-vton-latest`, in the US region only, and is billed at
-2x the standard realtime rate for those models. Other models ignore the option (the SDK emits a
-warning). Omit it (the default) for standard mode.
+`lucy-latest`, `lucy-vton-3.5`, and `lucy-vton-3.6` / `lucy-vton-latest` (`lucy-vton-latest`
+currently resolves to `lucy-vton-3.6`), in the US region only, and is billed at 2x the standard
+realtime rate for those models. Other models ignore the option (the SDK emits a warning). Omit
+it (the default) for standard mode.
 
 ```python
 from decart import DecartClient, models

decart/models.py

@@ -85,8 +85,9 @@ def _warn_unsupported_speed(model: "ModelDefinition[Any]", speed: str) -> None:
     _warned_unsupported_speeds.add(key)
     warnings.warn(
         f'Model "{model.name}" does not support speed="{speed}"; the option is ignored by the '
-        "server for this model. Fast mode is currently available for lucy-2.5 / lucy-latest and "
-        "lucy-vton-3.5 / lucy-vton-latest only. See https://docs.platform.decart.ai/models "
+        "server for this model. Fast mode is currently available for lucy-2.5 / lucy-latest, "
+        "lucy-vton-3.5, and lucy-vton-3.6 / lucy-vton-latest only. "
+        "See https://docs.platform.decart.ai/models "
         "for details.",
         UserWarning,
         stacklevel=3,
@@ -231,7 +232,7 @@ class ImageToImageInput(DecartBaseModel):
             height=624,
             supported_speeds=("fast",),
         ),
-        # Server-side alias currently resolves to lucy-vton-3.5.
+        # Server-side alias currently resolves to lucy-vton-3.6.
         "lucy-vton-latest": ModelDefinition(
             name="lucy-vton-latest",
             url_path="/v1/stream",
@@ -254,6 +255,7 @@ class ImageToImageInput(DecartBaseModel):
             fps=30,
             width=1280,
             height=720,
+            supported_speeds=("fast",),
         ),
         "lucy-restyle-latest": ModelDefinition(
             name="lucy-restyle-latest",
@@ -322,7 +324,7 @@ class ImageToImageInput(DecartBaseModel):
             height=624,
             input_schema=VideoEdit2Input,
         ),
-        # Server-side alias currently resolves to lucy-vton-3.5.
+        # Server-side alias currently resolves to lucy-vton-3.6.
         "lucy-vton-latest": ModelDefinition(
             name="lucy-vton-latest",
             url_path="/v1/jobs/lucy-vton-latest",

decart/realtime/types.py

@@ -23,6 +23,6 @@ class RealtimeConnectOptions:
     speed: Optional[RealtimeSpeed] = None
     """Realtime speed tier. ``"fast"`` serves the session from a higher-compute tier for
     lower latency and higher throughput; output quality is unchanged. Currently available
-    for lucy-2.5 / lucy-latest and lucy-vton-3.5 / lucy-vton-latest, in the US region only,
-    and billed at 2x the standard realtime rate for those models. Other models ignore the
-    option. Omit it (the default) for standard mode."""
+    for lucy-2.5 / lucy-latest, lucy-vton-3.5, and lucy-vton-3.6 / lucy-vton-latest, in the US
+    region only, and billed at 2x the standard realtime rate for those models. Other models
+    ignore the option. Omit it (the default) for standard mode."""

examples/README.md

@@ -37,9 +37,10 @@ pip install decart[realtime]
 
 Fast mode: pass `speed="fast"` on `RealtimeConnectOptions` to serve the session from a
 higher-compute tier for lower latency and higher throughput (output quality is unchanged). It is
-currently available for `lucy-2.5` / `lucy-latest` and `lucy-vton-3.5` / `lucy-vton-latest`, in
-the US region only, and is billed at 2x the standard realtime rate for those models. Other models
-ignore the option. Omit it (the default) for standard mode. The playground exposes it as
+currently available for `lucy-2.5` / `lucy-latest`, `lucy-vton-3.5`, and `lucy-vton-3.6` /
+`lucy-vton-latest`, in the US region only, and is billed at 2x the standard realtime rate for
+those models. Other models ignore the option. Omit it (the default) for standard mode. The
+playground exposes it as
 `python playground/playground.py --model lucy-2.5 --speed fast`.
 
 ### Running Examples

examples/realtime_synthetic.py

@@ -128,8 +128,9 @@ def on_error(error):
                         image=Path("examples/files/image.png"),
                     ),
                     # Fast mode: uncomment with a supporting model (lucy-2.5 / lucy-latest,
-                    # lucy-vton-3.5 / lucy-vton-latest). US region only, billed at 2x the
-                    # standard realtime rate; other models ignore it. Omit for standard mode.
+                    # lucy-vton-3.5, lucy-vton-3.6 / lucy-vton-latest). US region only, billed
+                    # at 2x the standard realtime rate; other models ignore it. Omit for
+                    # standard mode.
                     # speed="fast",
                 ),
             )

examples/video_tryon.py

@@ -1,8 +1,9 @@
 """
 Virtual Try-On Example
 
-This example demonstrates how to use the lucy-vton-latest model to perform
-virtual try-on on a video using a reference garment image.
+This example demonstrates how to use the lucy-vton-latest model (currently
+lucy-vton-3.6, resolved server-side) to perform virtual try-on on a video
+using a reference garment image.
 
 Usage:
     # With reference image and prompt:

playground/playground.py

@@ -125,8 +125,8 @@ def parse_args() -> argparse.Namespace:
         choices=["fast"],
         help=(
             "Realtime speed tier. 'fast' uses a higher-compute tier for lower latency "
-            "(lucy-2.5 / lucy-vton-3.5 and their -latest aliases only, US region, billed at 2x). "
-            "Omit for standard mode."
+            "(lucy-2.5, lucy-vton-3.5, lucy-vton-3.6 and their -latest aliases only, US region, "
+            "billed at 2x). Omit for standard mode."
         ),
     )
     parser.add_argument("--camera", "-c", type=int, default=0, help="Camera device index")

tests/test_models.py

@@ -38,8 +38,8 @@ def test_canonical_realtime_models() -> None:
     assert model.fps == 30
     assert model.width == 1280
     assert model.height == 720
-    # 3.6 is not offered on the fast tier, unlike 3.5.
-    assert model.supported_speeds == ()
+    # 3.6 is served on the fast tier, like 3.5.
+    assert model.supported_speeds == ("fast",)
 
 
 def test_canonical_video_models() -> None:
@@ -136,6 +136,7 @@ def test_latest_realtime_models() -> None:
     assert model.width == 1088
     assert model.height == 624
 
+    # Server-side alias currently resolves to lucy-vton-3.6 (same geometry).
     model = models.realtime("lucy-vton-latest")
     assert model.name == "lucy-vton-latest"
     assert model.url_path == "/v1/stream"
@@ -159,6 +160,7 @@ def test_latest_video_models() -> None:
     assert model.width == 1088
     assert model.height == 624
 
+    # Server-side alias currently resolves to lucy-vton-3.6 (same geometry).
     model = models.video("lucy-vton-latest")
     assert model.name == "lucy-vton-latest"
     assert model.url_path == "/v1/jobs/lucy-vton-latest"
@@ -199,10 +201,16 @@ def test_latest_aliases_no_deprecation_warning() -> None:
         assert len(w) == 0
 
 
-FAST_REALTIME_MODELS = {"lucy-2.5", "lucy-latest", "lucy-vton-3.5", "lucy-vton-latest"}
+FAST_REALTIME_MODELS = {
+    "lucy-2.5",
+    "lucy-latest",
+    "lucy-vton-3.5",
+    "lucy-vton-3.6",
+    "lucy-vton-latest",
+}
 
 
-def test_realtime_fast_speed_capability_pinned_to_lucy_2_5_and_vton_3_5() -> None:
+def test_realtime_fast_speed_capability_pinned_to_lucy_2_5_and_vton_3_5_3_6() -> None:
     # Matches the JS SDK registry: only these realtime entries advertise speed="fast".
     fast_models = {
         name for name, model in _MODELS["realtime"].items() if "fast" in model.supported_speeds

tests/test_realtime_unit.py

@@ -230,7 +230,13 @@ async def test_realtime_connect_speed_fast_no_warning_for_supported_model():
     _warned_unsupported_speeds.clear()
     with warnings.catch_warnings(record=True) as w:
         warnings.simplefilter("always")
-        for model in ("lucy-2.5", "lucy-latest", "lucy-vton-3.5", "lucy-vton-latest"):
+        for model in (
+            "lucy-2.5",
+            "lucy-latest",
+            "lucy-vton-3.5",
+            "lucy-vton-3.6",
+            "lucy-vton-latest",
+        ):
             url = await _connect_and_capture_url(speed="fast", model=model)
             assert "&speed=fast" in url
         assert _speed_warnings(w) == []