Merge remote-tracking branch 'origin/next' into sam/run-helper

Author: dtmeadows

.release-please-manifest.json

@@ -1,3 +1,3 @@
 {
-  ".": "0.1.0"
+  ".": "0.2.1"
 }

.stats.yml

@@ -1,4 +1,4 @@
 configured_endpoints: 35
-openapi_spec_url: https://storage.googleapis.com/stainless-sdk-openapi-specs/replicate%2Freplicate-client-efbc8cc2d74644b213e161d3e11e0589d1cef181fb318ea02c8eb6b00f245713.yml
-openapi_spec_hash: 13da0c06c900b61cd98ab678e024987a
-config_hash: 8ef6787524fd12bfeb27f8c6acef3dca
+openapi_spec_url: https://storage.googleapis.com/stainless-sdk-openapi-specs/replicate%2Freplicate-client-88cf5fe1f5accb56ae9fbb31c0df00d1552762d4c558d16d8547894ae95e8ccb.yml
+openapi_spec_hash: 43283d20f335a04241cce165452ff50e
+config_hash: 84794ed69d841684ff08a8aa889ef103

CHANGELOG.md

@@ -1,5 +1,27 @@
 # Changelog
 
+## 0.2.1 (2025-05-07)
+
+Full Changelog: [v0.2.0...v0.2.1](https://github.com/replicate/replicate-python-stainless/compare/v0.2.0...v0.2.1)
+
+### Documentation
+
+* update example requests ([eb0ba44](https://github.com/replicate/replicate-python-stainless/commit/eb0ba44af5b5006e758c9d9e65312f88b52dc3f5))
+
+## 0.2.0 (2025-05-07)
+
+Full Changelog: [v0.1.0...v0.2.0](https://github.com/replicate/replicate-python-stainless/compare/v0.1.0...v0.2.0)
+
+### Features
+
+* **api:** add Files API methods ([3173e5f](https://github.com/replicate/replicate-python-stainless/commit/3173e5f61edd89ffe0b64b53fc8e8e9905e145e4))
+* **api:** fix bearer token which also regressed when guessing with AI ([13162be](https://github.com/replicate/replicate-python-stainless/commit/13162be9d367de29d222b86506fa921a10800665))
+
+
+### Bug Fixes
+
+* **api:** fix client_settings.opts.api_key.read_env ([5a9b95c](https://github.com/replicate/replicate-python-stainless/commit/5a9b95ce89e536b539eefe0864a47784fdb0ec08))
+
 ## 0.1.0 (2025-05-07)
 
 Full Changelog: [v0.1.0-alpha.10...v0.1.0](https://github.com/replicate/replicate-python-stainless/compare/v0.1.0-alpha.10...v0.1.0)

README.md

@@ -31,8 +31,10 @@ client = Replicate(
     bearer_token=os.environ.get("REPLICATE_API_TOKEN"),  # This is the default and can be omitted
 )
 
-account = client.account.get()
-print(account.type)
+prediction = client.predictions.get(
+    prediction_id="gm3qorzdhgbfurvjtvhg6dckhu",
+)
+print(prediction.id)
 ```
 
 While you can provide a `bearer_token` keyword argument,
@@ -55,8 +57,10 @@ client = AsyncReplicate(
 
 
 async def main() -> None:
-    account = await client.account.get()
-    print(account.type)
+    prediction = await client.predictions.get(
+        prediction_id="gm3qorzdhgbfurvjtvhg6dckhu",
+    )
+    print(prediction.id)
 
 
 asyncio.run(main())
@@ -84,12 +88,12 @@ from replicate import Replicate
 
 client = Replicate()
 
-all_predictions = []
+all_models = []
 # Automatically fetches more pages as needed.
-for prediction in client.predictions.list():
-    # Do something with prediction here
-    all_predictions.append(prediction)
-print(all_predictions)
+for model in client.models.list():
+    # Do something with model here
+    all_models.append(model)
+print(all_models)
 ```
 
 Or, asynchronously:
@@ -102,11 +106,11 @@ client = AsyncReplicate()
 
 
 async def main() -> None:
-    all_predictions = []
+    all_models = []
     # Iterate through items across all pages, issuing requests as needed.
-    async for prediction in client.predictions.list():
-        all_predictions.append(prediction)
-    print(all_predictions)
+    async for model in client.models.list():
+        all_models.append(model)
+    print(all_models)
 
 
 asyncio.run(main())
@@ -115,7 +119,7 @@ asyncio.run(main())
 Alternatively, you can use the `.has_next_page()`, `.next_page_info()`, or `.get_next_page()` methods for more granular control working with pages:
 
 ```python
-first_page = await client.predictions.list()
+first_page = await client.models.list()
 if first_page.has_next_page():
     print(f"will fetch next page using these details: {first_page.next_page_info()}")
     next_page = await first_page.get_next_page()
@@ -127,11 +131,11 @@ if first_page.has_next_page():
 Or just work directly with the returned data:
 
 ```python
-first_page = await client.predictions.list()
+first_page = await client.models.list()
 
 print(f"next URL: {first_page.next}")  # => "next URL: ..."
-for prediction in first_page.results:
-    print(prediction.id)
+for model in first_page.results:
+    print(model.cover_image_url)
 
 # Remove `await` for non-async usage.
 ```
@@ -170,7 +174,10 @@ from replicate import Replicate
 client = Replicate()
 
 try:
-    client.account.get()
+    client.predictions.create(
+        input={"text": "Alice"},
+        version="replicate/hello-world:5c7d5dc6dd8bf75c1acaa8565735e7986bc5b66206b55cca93cb72c9bf15ccaa",
+    )
 except replicate.APIConnectionError as e:
     print("The server could not be reached")
     print(e.__cause__)  # an underlying Exception, likely raised within httpx.
@@ -213,7 +220,10 @@ client = Replicate(
 )
 
 # Or, configure per-request:
-client.with_options(max_retries=5).account.get()
+client.with_options(max_retries=5).predictions.create(
+    input={"text": "Alice"},
+    version="replicate/hello-world:5c7d5dc6dd8bf75c1acaa8565735e7986bc5b66206b55cca93cb72c9bf15ccaa",
+)
 ```
 
 ### Timeouts
@@ -236,7 +246,10 @@ client = Replicate(
 )
 
 # Override per-request:
-client.with_options(timeout=5.0).account.get()
+client.with_options(timeout=5.0).predictions.create(
+    input={"text": "Alice"},
+    version="replicate/hello-world:5c7d5dc6dd8bf75c1acaa8565735e7986bc5b66206b55cca93cb72c9bf15ccaa",
+)
 ```
 
 On timeout, an `APITimeoutError` is thrown.
@@ -277,11 +290,16 @@ The "raw" Response object can be accessed by prefixing `.with_raw_response.` to
 from replicate import Replicate
 
 client = Replicate()
-response = client.account.with_raw_response.get()
+response = client.predictions.with_raw_response.create(
+    input={
+        "text": "Alice"
+    },
+    version="replicate/hello-world:5c7d5dc6dd8bf75c1acaa8565735e7986bc5b66206b55cca93cb72c9bf15ccaa",
+)
 print(response.headers.get('X-My-Header'))
 
-account = response.parse()  # get the object that `account.get()` would have returned
-print(account.type)
+prediction = response.parse()  # get the object that `predictions.create()` would have returned
+print(prediction.id)
 ```
 
 These methods return an [`APIResponse`](https://github.com/replicate/replicate-python-stainless/tree/main/src/replicate/_response.py) object.
@@ -295,7 +313,10 @@ The above interface eagerly reads the full response body when you make the reque
 To stream the response body, use `.with_streaming_response` instead, which requires a context manager and only reads the response body once you call `.read()`, `.text()`, `.json()`, `.iter_bytes()`, `.iter_text()`, `.iter_lines()` or `.parse()`. In the async client, these are async methods.
 
 ```python
-with client.account.with_streaming_response.get() as response:
+with client.predictions.with_streaming_response.create(
+    input={"text": "Alice"},
+    version="replicate/hello-world:5c7d5dc6dd8bf75c1acaa8565735e7986bc5b66206b55cca93cb72c9bf15ccaa",
+) as response:
     print(response.headers.get("X-My-Header"))
 
     for line in response.iter_lines():

api.md

@@ -116,7 +116,7 @@ Methods:
 
 - <code title="post /predictions">client.predictions.<a href="./src/replicate/resources/predictions.py">create</a>(\*\*<a href="src/replicate/types/prediction_create_params.py">params</a>) -> <a href="./src/replicate/types/prediction.py">Prediction</a></code>
 - <code title="get /predictions">client.predictions.<a href="./src/replicate/resources/predictions.py">list</a>(\*\*<a href="src/replicate/types/prediction_list_params.py">params</a>) -> <a href="./src/replicate/types/prediction.py">SyncCursorURLPageWithCreatedFilters[Prediction]</a></code>
-- <code title="post /predictions/{prediction_id}/cancel">client.predictions.<a href="./src/replicate/resources/predictions.py">cancel</a>(\*, prediction_id) -> None</code>
+- <code title="post /predictions/{prediction_id}/cancel">client.predictions.<a href="./src/replicate/resources/predictions.py">cancel</a>(\*, prediction_id) -> <a href="./src/replicate/types/prediction.py">Prediction</a></code>
 - <code title="get /predictions/{prediction_id}">client.predictions.<a href="./src/replicate/resources/predictions.py">get</a>(\*, prediction_id) -> <a href="./src/replicate/types/prediction.py">Prediction</a></code>
 
 # Trainings

pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "replicate-stainless"
-version = "0.1.0"
+version = "0.2.1"
 description = "The official Python library for the replicate API"
 dynamic = ["readme"]
 license = "Apache-2.0"

src/replicate/_version.py

@@ -1,4 +1,4 @@
 # File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details.
 
 __title__ = "replicate"
-__version__ = "0.1.0"  # x-release-please-version
+__version__ = "0.2.1"  # x-release-please-version

src/replicate/resources/predictions.py

@@ -11,7 +11,7 @@
 from replicate.lib._files import FileEncodingStrategy, encode_json, async_encode_json
 
 from ..types import prediction_list_params, prediction_create_params
-from .._types import NOT_GIVEN, Body, Query, Headers, NoneType, NotGiven
+from .._types import NOT_GIVEN, Body, Query, Headers, NotGiven
 from .._utils import maybe_transform, strip_not_given, async_maybe_transform
 from .._compat import cached_property
 from .._resource import SyncAPIResource, AsyncAPIResource
@@ -330,9 +330,32 @@ def cancel(
         extra_query: Query | None = None,
         extra_body: Body | None = None,
         timeout: float | httpx.Timeout | None | NotGiven = NOT_GIVEN,
-    ) -> None:
+    ) -> Prediction:
         """
-        Cancel a prediction
+        Cancel a prediction that is currently running.
+
+        Example cURL request that creates a prediction and then cancels it:
+
+        ```console
+        # First, create a prediction
+        PREDICTION_ID=$(curl -s -X POST \\
+          -H "Authorization: Bearer $REPLICATE_API_TOKEN" \\
+          -H "Content-Type: application/json" \\
+          -d '{
+            "input": {
+              "prompt": "a video that may take a while to generate"
+            }
+          }' \\
+          https://api.replicate.com/v1/models/minimax/video-01/predictions | jq -r '.id')
+
+        # Echo the prediction ID
+        echo "Created prediction with ID: $PREDICTION_ID"
+
+        # Cancel the prediction
+        curl -s -X POST \\
+          -H "Authorization: Bearer $REPLICATE_API_TOKEN" \\
+          https://api.replicate.com/v1/predictions/$PREDICTION_ID/cancel
+        ```
 
         Args:
           extra_headers: Send extra headers
@@ -345,13 +368,12 @@ def cancel(
         """
         if not prediction_id:
             raise ValueError(f"Expected a non-empty value for `prediction_id` but received {prediction_id!r}")
-        extra_headers = {"Accept": "*/*", **(extra_headers or {})}
         return self._post(
             f"/predictions/{prediction_id}/cancel",
             options=make_request_options(
                 extra_headers=extra_headers, extra_query=extra_query, extra_body=extra_body, timeout=timeout
             ),
-            cast_to=NoneType,
+            cast_to=Prediction,
         )
 
     def get(
@@ -759,9 +781,32 @@ async def cancel(
         extra_query: Query | None = None,
         extra_body: Body | None = None,
         timeout: float | httpx.Timeout | None | NotGiven = NOT_GIVEN,
-    ) -> None:
+    ) -> Prediction:
         """
-        Cancel a prediction
+        Cancel a prediction that is currently running.
+
+        Example cURL request that creates a prediction and then cancels it:
+
+        ```console
+        # First, create a prediction
+        PREDICTION_ID=$(curl -s -X POST \\
+          -H "Authorization: Bearer $REPLICATE_API_TOKEN" \\
+          -H "Content-Type: application/json" \\
+          -d '{
+            "input": {
+              "prompt": "a video that may take a while to generate"
+            }
+          }' \\
+          https://api.replicate.com/v1/models/minimax/video-01/predictions | jq -r '.id')
+
+        # Echo the prediction ID
+        echo "Created prediction with ID: $PREDICTION_ID"
+
+        # Cancel the prediction
+        curl -s -X POST \\
+          -H "Authorization: Bearer $REPLICATE_API_TOKEN" \\
+          https://api.replicate.com/v1/predictions/$PREDICTION_ID/cancel
+        ```
 
         Args:
           extra_headers: Send extra headers
@@ -774,13 +819,12 @@ async def cancel(
         """
         if not prediction_id:
             raise ValueError(f"Expected a non-empty value for `prediction_id` but received {prediction_id!r}")
-        extra_headers = {"Accept": "*/*", **(extra_headers or {})}
         return await self._post(
             f"/predictions/{prediction_id}/cancel",
             options=make_request_options(
                 extra_headers=extra_headers, extra_query=extra_query, extra_body=extra_body, timeout=timeout
             ),
-            cast_to=NoneType,
+            cast_to=Prediction,
         )
 
     async def get(

tests/api_resources/test_predictions.py

@@ -112,7 +112,7 @@ def test_method_cancel(self, client: Replicate) -> None:
         prediction = client.predictions.cancel(
             prediction_id="prediction_id",
         )
-        assert prediction is None
+        assert_matches_type(Prediction, prediction, path=["response"])
 
     @pytest.mark.skip()
     @parametrize
@@ -124,7 +124,7 @@ def test_raw_response_cancel(self, client: Replicate) -> None:
         assert response.is_closed is True
         assert response.http_request.headers.get("X-Stainless-Lang") == "python"
         prediction = response.parse()
-        assert prediction is None
+        assert_matches_type(Prediction, prediction, path=["response"])
 
     @pytest.mark.skip()
     @parametrize
@@ -136,7 +136,7 @@ def test_streaming_response_cancel(self, client: Replicate) -> None:
             assert response.http_request.headers.get("X-Stainless-Lang") == "python"
 
             prediction = response.parse()
-            assert prediction is None
+            assert_matches_type(Prediction, prediction, path=["response"])
 
         assert cast(Any, response.is_closed) is True
 
@@ -287,7 +287,7 @@ async def test_method_cancel(self, async_client: AsyncReplicate) -> None:
         prediction = await async_client.predictions.cancel(
             prediction_id="prediction_id",
         )
-        assert prediction is None
+        assert_matches_type(Prediction, prediction, path=["response"])
 
     @pytest.mark.skip()
     @parametrize
@@ -299,7 +299,7 @@ async def test_raw_response_cancel(self, async_client: AsyncReplicate) -> None:
         assert response.is_closed is True
         assert response.http_request.headers.get("X-Stainless-Lang") == "python"
         prediction = await response.parse()
-        assert prediction is None
+        assert_matches_type(Prediction, prediction, path=["response"])
 
     @pytest.mark.skip()
     @parametrize
@@ -311,7 +311,7 @@ async def test_streaming_response_cancel(self, async_client: AsyncReplicate) ->
             assert response.http_request.headers.get("X-Stainless-Lang") == "python"
 
             prediction = await response.parse()
-            assert prediction is None
+            assert_matches_type(Prediction, prediction, path=["response"])
 
         assert cast(Any, response.is_closed) is True