release: 2.0.0-alpha.2

.release-please-manifest.json

@@ -1,3 +1,3 @@
 {
-  ".": "2.0.0-alpha.1"
+  ".": "2.0.0-alpha.2"
 }

.stats.yml

@@ -1,4 +1,4 @@
 configured_endpoints: 35
 openapi_spec_url: https://storage.googleapis.com/stainless-sdk-openapi-specs/replicate%2Freplicate-client-37cd8ea847eb57706035f766ca549d5b4e2111053af0656a2df9a8150421428e.yml
 openapi_spec_hash: a3e4d6fd9aff6de0e4b6d8ad28cbbe05
-config_hash: da444f7a7ac6238fa0bdecaa01ffa4c3
+config_hash: 12536d2bf978a995771d076a4647c17d

CHANGELOG.md

@@ -1,5 +1,20 @@
 # Changelog
 
+## 2.0.0-alpha.2 (2025-06-13)
+
+Full Changelog: [v2.0.0-alpha.1...v2.0.0-alpha.2](https://github.com/replicate/replicate-python-stainless/compare/v2.0.0-alpha.1...v2.0.0-alpha.2)
+
+### Bug Fixes
+
+* **client:** correctly parse binary response | stream ([ec19335](https://github.com/replicate/replicate-python-stainless/commit/ec19335fcce5a7ceba6aa1a4ac67411421a571ec))
+* **test:** update prediction response ([b6608ca](https://github.com/replicate/replicate-python-stainless/commit/b6608ca5bbfc38d68a9bbfb853bbb8645e046d39))
+
+
+### Chores
+
+* **internal:** version bump ([497fe10](https://github.com/replicate/replicate-python-stainless/commit/497fe1047ade040d99bf5d7d4fb99f1b1b95a09d))
+* **tests:** run tests in parallel ([9330c56](https://github.com/replicate/replicate-python-stainless/commit/9330c565de11b0c61e43df899f4113c5214572c9))
+
 ## 2.0.0-alpha.1 (2025-06-10)
 
 Full Changelog: [v0.6.0...v2.0.0-alpha.1](https://github.com/replicate/replicate-python-stainless/compare/v0.6.0...v2.0.0-alpha.1)

pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "replicate"
-version = "2.0.0-alpha.1"
+version = "2.0.0-alpha.2"
 description = "The official Python library for the replicate API"
 dynamic = ["readme"]
 license = "Apache-2.0"
@@ -54,6 +54,7 @@ dev-dependencies = [
     "importlib-metadata>=6.7.0",
     "rich>=13.7.1",
     "nest_asyncio==1.6.0",
+    "pytest-xdist>=3.6.1",
 ]
 
 [tool.rye.scripts]
@@ -125,7 +126,7 @@ replacement = '[\1](https://github.com/replicate/replicate-python-stainless/tree
 
 [tool.pytest.ini_options]
 testpaths = ["tests"]
-addopts = "--tb=short"
+addopts = "--tb=short -n auto"
 xfail_strict = true
 asyncio_mode = "auto"
 asyncio_default_fixture_loop_scope = "session"

requirements-dev.lock

@@ -30,6 +30,8 @@ distro==1.8.0
 exceptiongroup==1.2.2
     # via anyio
     # via pytest
+execnet==2.1.1
+    # via pytest-xdist
 filelock==3.12.4
     # via virtualenv
 h11==0.14.0
@@ -72,7 +74,9 @@ pygments==2.18.0
 pyright==1.1.399
 pytest==8.3.3
     # via pytest-asyncio
+    # via pytest-xdist
 pytest-asyncio==0.24.0
+pytest-xdist==3.7.0
 python-dateutil==2.8.2
     # via time-machine
 pytz==2023.3.post1

scripts/utils/upload-artifact.sh

@@ -18,7 +18,7 @@ UPLOAD_RESPONSE=$(tar -cz . | curl -v -X PUT \
 
 if echo "$UPLOAD_RESPONSE" | grep -q "HTTP/[0-9.]* 200"; then
   echo -e "\033[32mUploaded build to Stainless storage.\033[0m"
-  echo -e "\033[32mInstallation: pip install 'https://pkg.stainless.com/s/replicate-client-python/$SHA'\033[0m"
+  echo -e "\033[32mInstallation: pip install --pre 'https://pkg.stainless.com/s/replicate-client-python/$SHA'\033[0m"
 else
   echo -e "\033[31mFailed to upload artifact.\033[0m"
   exit 1

src/replicate/_base_client.py

@@ -1071,7 +1071,14 @@ def _process_response(
     ) -> ResponseT:
         origin = get_origin(cast_to) or cast_to
 
-        if inspect.isclass(origin) and issubclass(origin, BaseAPIResponse):
+        if (
+            inspect.isclass(origin)
+            and issubclass(origin, BaseAPIResponse)
+            # we only want to actually return the custom BaseAPIResponse class if we're
+            # returning the raw response, or if we're not streaming SSE, as if we're streaming
+            # SSE then `cast_to` doesn't actively reflect the type we need to parse into
+            and (not stream or bool(response.request.headers.get(RAW_RESPONSE_HEADER)))
+        ):
             if not issubclass(origin, APIResponse):
                 raise TypeError(f"API Response types must subclass {APIResponse}; Received {origin}")
 
@@ -1588,7 +1595,14 @@ async def _process_response(
     ) -> ResponseT:
         origin = get_origin(cast_to) or cast_to
 
-        if inspect.isclass(origin) and issubclass(origin, BaseAPIResponse):
+        if (
+            inspect.isclass(origin)
+            and issubclass(origin, BaseAPIResponse)
+            # we only want to actually return the custom BaseAPIResponse class if we're
+            # returning the raw response, or if we're not streaming SSE, as if we're streaming
+            # SSE then `cast_to` doesn't actively reflect the type we need to parse into
+            and (not stream or bool(response.request.headers.get(RAW_RESPONSE_HEADER)))
+        ):
             if not issubclass(origin, AsyncAPIResponse):
                 raise TypeError(f"API Response types must subclass {AsyncAPIResponse}; Received {origin}")
 

src/replicate/_version.py

@@ -1,4 +1,4 @@
 # File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details.
 
 __title__ = "replicate"
-__version__ = "2.0.0-alpha.1"  # x-release-please-version
+__version__ = "2.0.0-alpha.2"  # x-release-please-version

tests/lib/test_run.py

@@ -32,6 +32,7 @@ def create_mock_prediction(
         urls = {
             "get": "https://api.replicate.com/v1/predictions/test_prediction_id",
             "cancel": "https://api.replicate.com/v1/predictions/test_prediction_id/cancel",
+            "web": "https://replicate.com/p/test_prediction_id",
         }
 
     return {