release: 1.83.0 (#2393)

* Fix a typo: "occurences" -> "occurrences" (#2387)

* chore: deprecate Assistants API

* chore(api): mark some methods as deprecated

* feat(api): Config update for pakrym-stream-param

* fix(client): return binary content from `get /containers/{container_id}/files/{file_id}/content`

* codegen metadata

* chore(docs): remove reference to rye shell

* feat(client): add follow_redirects request option

* fix(api): Fix evals and code interpreter interfaces

* release: 1.83.0

---------

Co-authored-by: Roman A <[email protected]>
Co-authored-by: stainless-app[bot] <142633134+stainless-app[bot]@users.noreply.github.com>

Author: stainless-app[bot]

.release-please-manifest.json

@@ -1,3 +1,3 @@
 {
-  ".": "1.82.1"
+  ".": "1.83.0"
 }

.stats.yml

@@ -1,4 +1,4 @@
 configured_endpoints: 111
-openapi_spec_url: https://storage.googleapis.com/stainless-sdk-openapi-specs/openai%2Fopenai-fc64d7c2c8f51f750813375356c3f3fdfc7fc1b1b34f19c20a5410279d445d37.yml
-openapi_spec_hash: 618285fc70199ee32b9ebe4bf72f7e4c
-config_hash: 535b6e5f26a295d609b259c8cb8f656c
+openapi_spec_url: https://storage.googleapis.com/stainless-sdk-openapi-specs/openai%2Fopenai-2bcc845d8635bf93ddcf9ee723af4d7928248412a417bee5fc10d863a1e13867.yml
+openapi_spec_hash: 865230cb3abeb01bd85de05891af23c4
+config_hash: ed1e6b3c5f93d12b80d31167f55c557c

CHANGELOG.md

@@ -1,5 +1,27 @@
 # Changelog
 
+## 1.83.0 (2025-06-02)
+
+Full Changelog: [v1.82.1...v1.83.0](https://github.com/openai/openai-python/compare/v1.82.1...v1.83.0)
+
+### Features
+
+* **api:** Config update for pakrym-stream-param ([88bcf3a](https://github.com/openai/openai-python/commit/88bcf3af9ce8ffa8347547d4d30aacac1ceba939))
+* **client:** add follow_redirects request option ([26d715f](https://github.com/openai/openai-python/commit/26d715f4e9b0f2b19e2ac16acc796a949338e1e1))
+
+
+### Bug Fixes
+
+* **api:** Fix evals and code interpreter interfaces ([2650159](https://github.com/openai/openai-python/commit/2650159f6d01f6eb481cf8c7942142e4fd21ce44))
+* **client:** return binary content from `get /containers/{container_id}/files/{file_id}/content` ([f7c80c4](https://github.com/openai/openai-python/commit/f7c80c4368434bd0be7436375076ba33a62f63b5))
+
+
+### Chores
+
+* **api:** mark some methods as deprecated ([3e2ca57](https://github.com/openai/openai-python/commit/3e2ca571cb6cdd9e15596590605b2f98a4c5a42e))
+* deprecate Assistants API ([9d166d7](https://github.com/openai/openai-python/commit/9d166d795e03dea49af680ec9597e9497522187c))
+* **docs:** remove reference to rye shell ([c7978e9](https://github.com/openai/openai-python/commit/c7978e9f1640c311022988fcd716cbb5c865daa8))
+
 ## 1.82.1 (2025-05-29)
 
 Full Changelog: [v1.82.0...v1.82.1](https://github.com/openai/openai-python/compare/v1.82.0...v1.82.1)

CONTRIBUTING.md

@@ -17,8 +17,7 @@ $ rye sync --all-features
 You can then run scripts using `rye run python script.py` or by activating the virtual environment:
 
 ```sh
-$ rye shell
-# or manually activate - https://docs.python.org/3/library/venv.html#how-venvs-work
+# Activate the virtual environment - https://docs.python.org/3/library/venv.html#how-venvs-work
 $ source .venv/bin/activate
 
 # now you can omit the `rye run` prefix

api.md

@@ -784,7 +784,7 @@ Methods:
 - <code title="post /responses">client.responses.<a href="./src/openai/resources/responses/responses.py">create</a>(\*\*<a href="src/openai/types/responses/response_create_params.py">params</a>) -> <a href="./src/openai/types/responses/response.py">Response</a></code>
 - <code title="get /responses/{response_id}">client.responses.<a href="./src/openai/resources/responses/responses.py">retrieve</a>(response_id, \*\*<a href="src/openai/types/responses/response_retrieve_params.py">params</a>) -> <a href="./src/openai/types/responses/response.py">Response</a></code>
 - <code title="delete /responses/{response_id}">client.responses.<a href="./src/openai/resources/responses/responses.py">delete</a>(response_id) -> None</code>
-- <code title="post /responses/{response_id}/cancel">client.responses.<a href="./src/openai/resources/responses/responses.py">cancel</a>(response_id) -> None</code>
+- <code title="post /responses/{response_id}/cancel">client.responses.<a href="./src/openai/resources/responses/responses.py">cancel</a>(response_id) -> <a href="./src/openai/types/responses/response.py">Response</a></code>
 
 ## InputItems
 
@@ -894,4 +894,4 @@ Methods:
 
 Methods:
 
-- <code title="get /containers/{container_id}/files/{file_id}/content">client.containers.files.content.<a href="./src/openai/resources/containers/files/content.py">retrieve</a>(file_id, \*, container_id) -> None</code>
+- <code title="get /containers/{container_id}/files/{file_id}/content">client.containers.files.content.<a href="./src/openai/resources/containers/files/content.py">retrieve</a>(file_id, \*, container_id) -> HttpxBinaryResponseContent</code>

examples/assistant.py

@@ -1,6 +1,6 @@
 [project]
 name = "openai"
-version = "1.82.1"
+version = "1.83.0"
 description = "The official Python library for the openai API"
 dynamic = ["readme"]
 license = "Apache-2.0"

examples/assistant_stream.py

@@ -962,6 +962,9 @@ def request(
             if self.custom_auth is not None:
                 kwargs["auth"] = self.custom_auth
 
+            if options.follow_redirects is not None:
+                kwargs["follow_redirects"] = options.follow_redirects
+
             log.debug("Sending HTTP Request: %s %s", request.method, request.url)
 
             response = None
@@ -1477,6 +1480,9 @@ async def request(
             if self.custom_auth is not None:
                 kwargs["auth"] = self.custom_auth
 
+            if options.follow_redirects is not None:
+                kwargs["follow_redirects"] = options.follow_redirects
+
             log.debug("Sending HTTP Request: %s %s", request.method, request.url)
 
             response = None

examples/assistant_stream_helpers.py

@@ -777,6 +777,7 @@ class FinalRequestOptionsInput(TypedDict, total=False):
     idempotency_key: str
     json_data: Body
     extra_json: AnyMapping
+    follow_redirects: bool
 
 
 @final
@@ -790,6 +791,7 @@ class FinalRequestOptions(pydantic.BaseModel):
     files: Union[HttpxRequestFiles, None] = None
     idempotency_key: Union[str, None] = None
     post_parser: Union[Callable[[Any], Any], NotGiven] = NotGiven()
+    follow_redirects: Union[bool, None] = None
 
     # It should be noted that we cannot use `json` here as that would override
     # a BaseModel method in an incompatible fashion.

pyproject.toml

@@ -101,6 +101,7 @@ class RequestOptions(TypedDict, total=False):
     params: Query
     extra_json: AnyMapping
     idempotency_key: str
+    follow_redirects: bool
 
 
 # Sentinel class used until PEP 0661 is accepted
@@ -217,3 +218,4 @@ class _GenericAlias(Protocol):
 
 class HttpxSendArgs(TypedDict, total=False):
     auth: httpx.Auth
+    follow_redirects: bool

src/openai/_base_client.py

@@ -212,7 +212,7 @@ def _transform_recursive(
         return data
 
     if isinstance(data, pydantic.BaseModel):
-        return model_dump(data, exclude_unset=True, mode="json", exclude=getattr(data, '__api_exclude__', None))
+        return model_dump(data, exclude_unset=True, mode="json", exclude=getattr(data, "__api_exclude__", None))
 
     annotated_type = _get_annotated_type(annotation)
     if annotated_type is None:

src/openai/_models.py

@@ -1,4 +1,4 @@
 # File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details.
 
 __title__ = "openai"
-__version__ = "1.82.1"  # x-release-please-version
+__version__ = "1.83.0"  # x-release-please-version

src/openai/_types.py

@@ -109,7 +109,7 @@ def parse_response(
             or output.type == "code_interpreter_call"
             or output.type == "local_shell_call"
             or output.type == "mcp_list_tools"
-            or output.type == 'exec'
+            or output.type == "exec"
         ):
             output_list.append(output)
         elif TYPE_CHECKING:  # type: ignore

src/openai/_utils/_transform.py

@@ -43,6 +43,7 @@ def with_streaming_response(self) -> SessionsWithStreamingResponse:
     def create(
         self,
         *,
+        client_secret: session_create_params.ClientSecret | NotGiven = NOT_GIVEN,
         input_audio_format: Literal["pcm16", "g711_ulaw", "g711_alaw"] | NotGiven = NOT_GIVEN,
         input_audio_noise_reduction: session_create_params.InputAudioNoiseReduction | NotGiven = NOT_GIVEN,
         input_audio_transcription: session_create_params.InputAudioTranscription | NotGiven = NOT_GIVEN,
@@ -83,6 +84,8 @@ def create(
         the Realtime API.
 
         Args:
+          client_secret: Configuration options for the generated client secret.
+
           input_audio_format: The format of input audio. Options are `pcm16`, `g711_ulaw`, or `g711_alaw`. For
               `pcm16`, input audio must be 16-bit PCM at a 24kHz sample rate, single channel
               (mono), and little-endian byte order.
@@ -163,6 +166,7 @@ def create(
             "/realtime/sessions",
             body=maybe_transform(
                 {
+                    "client_secret": client_secret,
                     "input_audio_format": input_audio_format,
                     "input_audio_noise_reduction": input_audio_noise_reduction,
                     "input_audio_transcription": input_audio_transcription,
@@ -209,6 +213,7 @@ def with_streaming_response(self) -> AsyncSessionsWithStreamingResponse:
     async def create(
         self,
         *,
+        client_secret: session_create_params.ClientSecret | NotGiven = NOT_GIVEN,
         input_audio_format: Literal["pcm16", "g711_ulaw", "g711_alaw"] | NotGiven = NOT_GIVEN,
         input_audio_noise_reduction: session_create_params.InputAudioNoiseReduction | NotGiven = NOT_GIVEN,
         input_audio_transcription: session_create_params.InputAudioTranscription | NotGiven = NOT_GIVEN,
@@ -249,6 +254,8 @@ async def create(
         the Realtime API.
 
         Args:
+          client_secret: Configuration options for the generated client secret.
+
           input_audio_format: The format of input audio. Options are `pcm16`, `g711_ulaw`, or `g711_alaw`. For
               `pcm16`, input audio must be 16-bit PCM at a 24kHz sample rate, single channel
               (mono), and little-endian byte order.
@@ -329,6 +336,7 @@ async def create(
             "/realtime/sessions",
             body=await async_maybe_transform(
                 {
+                    "client_secret": client_secret,
                     "input_audio_format": input_audio_format,
                     "input_audio_noise_reduction": input_audio_noise_reduction,
                     "input_audio_transcription": input_audio_transcription,

src/openai/_version.py

@@ -43,6 +43,7 @@ def with_streaming_response(self) -> TranscriptionSessionsWithStreamingResponse:
     def create(
         self,
         *,
+        client_secret: transcription_session_create_params.ClientSecret | NotGiven = NOT_GIVEN,
         include: List[str] | NotGiven = NOT_GIVEN,
         input_audio_format: Literal["pcm16", "g711_ulaw", "g711_alaw"] | NotGiven = NOT_GIVEN,
         input_audio_noise_reduction: transcription_session_create_params.InputAudioNoiseReduction
@@ -67,6 +68,8 @@ def create(
         the Realtime API.
 
         Args:
+          client_secret: Configuration options for the generated client secret.
+
           include:
               The set of items to include in the transcription. Current available items are:
 
@@ -113,6 +116,7 @@ def create(
             "/realtime/transcription_sessions",
             body=maybe_transform(
                 {
+                    "client_secret": client_secret,
                     "include": include,
                     "input_audio_format": input_audio_format,
                     "input_audio_noise_reduction": input_audio_noise_reduction,
@@ -152,6 +156,7 @@ def with_streaming_response(self) -> AsyncTranscriptionSessionsWithStreamingResp
     async def create(
         self,
         *,
+        client_secret: transcription_session_create_params.ClientSecret | NotGiven = NOT_GIVEN,
         include: List[str] | NotGiven = NOT_GIVEN,
         input_audio_format: Literal["pcm16", "g711_ulaw", "g711_alaw"] | NotGiven = NOT_GIVEN,
         input_audio_noise_reduction: transcription_session_create_params.InputAudioNoiseReduction
@@ -176,6 +181,8 @@ async def create(
         the Realtime API.
 
         Args:
+          client_secret: Configuration options for the generated client secret.
+
           include:
               The set of items to include in the transcription. Current available items are:
 
@@ -222,6 +229,7 @@ async def create(
             "/realtime/transcription_sessions",
             body=await async_maybe_transform(
                 {
+                    "client_secret": client_secret,
                     "include": include,
                     "input_audio_format": input_audio_format,
                     "input_audio_noise_reduction": input_audio_noise_reduction,