Merge branch 'main' into feature/interceptor

Author: Nisarg38

Diff for: .release-please-manifest.json

@@ -1,3 +1,3 @@
 {
-  ".": "1.59.6"
+  ".": "1.59.8"
 }

Diff for: CHANGELOG.md

@@ -1,5 +1,37 @@
 # Changelog
 
+## 1.59.8 (2025-01-17)
+
+Full Changelog: [v1.59.7...v1.59.8](https://github.com/openai/openai-python/compare/v1.59.7...v1.59.8)
+
+### Bug Fixes
+
+* streaming ([c16f58e](https://github.com/openai/openai-python/commit/c16f58ead0bc85055b164182689ba74b7e939dfa))
+* **structured outputs:** avoid parsing empty empty content ([#2023](https://github.com/openai/openai-python/issues/2023)) ([6d3513c](https://github.com/openai/openai-python/commit/6d3513c86f6e5800f8f73a45e089b7a205327121))
+* **structured outputs:** correct schema coercion for inline ref expansion ([#2025](https://github.com/openai/openai-python/issues/2025)) ([2f4f0b3](https://github.com/openai/openai-python/commit/2f4f0b374207f162060c328b71ec995049dc42e8))
+* **types:** correct type for vector store chunking strategy ([#2017](https://github.com/openai/openai-python/issues/2017)) ([e389279](https://github.com/openai/openai-python/commit/e38927950a5cdad99065853fe7b72aad6bb322e9))
+
+
+### Chores
+
+* **examples:** update realtime model ([f26746c](https://github.com/openai/openai-python/commit/f26746cbcd893d66cf8a3fd68a7c3779dc8c833c)), closes [#2020](https://github.com/openai/openai-python/issues/2020)
+* **internal:** bump pyright dependency ([#2021](https://github.com/openai/openai-python/issues/2021)) ([0a9a0f5](https://github.com/openai/openai-python/commit/0a9a0f5d8b9d5457643798287f893305006dd518))
+* **internal:** streaming refactors ([#2012](https://github.com/openai/openai-python/issues/2012)) ([d76a748](https://github.com/openai/openai-python/commit/d76a748f606743407f94dfc26758095560e2082a))
+* **internal:** update deps ([#2015](https://github.com/openai/openai-python/issues/2015)) ([514e0e4](https://github.com/openai/openai-python/commit/514e0e415f87ab4510262d29ed6125384e017b84))
+
+
+### Documentation
+
+* **examples/azure:** example script with realtime API ([#1967](https://github.com/openai/openai-python/issues/1967)) ([84f2f9c](https://github.com/openai/openai-python/commit/84f2f9c0439229a7db7136fe78419292d34d1f81))
+
+## 1.59.7 (2025-01-13)
+
+Full Changelog: [v1.59.6...v1.59.7](https://github.com/openai/openai-python/compare/v1.59.6...v1.59.7)
+
+### Chores
+
+* export HttpxBinaryResponseContent class ([7191b71](https://github.com/openai/openai-python/commit/7191b71f3dcbbfcb2f2bec855c3bba93c956384e))
+
 ## 1.59.6 (2025-01-09)
 
 Full Changelog: [v1.59.5...v1.59.6](https://github.com/openai/openai-python/compare/v1.59.5...v1.59.6)

Diff for: README.md

@@ -275,7 +275,7 @@ from openai import AsyncOpenAI
 async def main():
     client = AsyncOpenAI()
 
-    async with client.beta.realtime.connect(model="gpt-4o-realtime-preview-2024-10-01") as connection:
+    async with client.beta.realtime.connect(model="gpt-4o-realtime-preview") as connection:
         await connection.session.update(session={'modalities': ['text']})
 
         await connection.conversation.item.create(
@@ -309,7 +309,7 @@ Whenever an error occurs, the Realtime API will send an [`error` event](https://
 ```py
 client = AsyncOpenAI()
 
-async with client.beta.realtime.connect(model="gpt-4o-realtime-preview-2024-10-01") as connection:
+async with client.beta.realtime.connect(model="gpt-4o-realtime-preview") as connection:
     ...
     async for event in connection:
         if event.type == 'error':

Diff for: api.md

@@ -314,7 +314,7 @@ from openai.types.beta import (
     OtherFileChunkingStrategyObject,
     StaticFileChunkingStrategy,
     StaticFileChunkingStrategyObject,
-    StaticFileChunkingStrategyParam,
+    StaticFileChunkingStrategyObjectParam,
     VectorStore,
     VectorStoreDeleted,
 )

Diff for: examples/realtime/azure_realtime.py

@@ -0,0 +1,57 @@
+import os
+import asyncio
+
+from azure.identity.aio import DefaultAzureCredential, get_bearer_token_provider
+
+from openai import AsyncAzureOpenAI
+
+# Azure OpenAI Realtime Docs
+
+# How-to: https://learn.microsoft.com/azure/ai-services/openai/how-to/realtime-audio
+# Supported models and API versions: https://learn.microsoft.com/azure/ai-services/openai/how-to/realtime-audio#supported-models
+# Entra ID auth: https://learn.microsoft.com/azure/ai-services/openai/how-to/managed-identity
+
+
+async def main() -> None:
+    """The following example demonstrates how to configure Azure OpenAI to use the Realtime API.
+    For an audio example, see push_to_talk_app.py and update the client and model parameter accordingly.
+
+    When prompted for user input, type a message and hit enter to send it to the model.
+    Enter "q" to quit the conversation.
+    """
+
+    credential = DefaultAzureCredential()
+    client = AsyncAzureOpenAI(
+        azure_endpoint=os.environ["AZURE_OPENAI_ENDPOINT"],
+        azure_ad_token_provider=get_bearer_token_provider(credential, "https://cognitiveservices.azure.com/.default"),
+        api_version="2024-10-01-preview",
+    )
+    async with client.beta.realtime.connect(
+        model="gpt-4o-realtime-preview",  # deployment name for your model
+    ) as connection:
+        await connection.session.update(session={"modalities": ["text"]})  # type: ignore
+        while True:
+            user_input = input("Enter a message: ")
+            if user_input == "q":
+                break
+
+            await connection.conversation.item.create(
+                item={
+                    "type": "message",
+                    "role": "user",
+                    "content": [{"type": "input_text", "text": user_input}],
+                }
+            )
+            await connection.response.create()
+            async for event in connection:
+                if event.type == "response.text.delta":
+                    print(event.delta, flush=True, end="")
+                elif event.type == "response.text.done":
+                    print()
+                elif event.type == "response.done":
+                    break
+
+    await credential.close()
+
+
+asyncio.run(main())

Diff for: examples/realtime/push_to_talk_app.py

@@ -152,7 +152,7 @@ async def on_mount(self) -> None:
         self.run_worker(self.send_mic_audio())
 
     async def handle_realtime_connection(self) -> None:
-        async with self.client.beta.realtime.connect(model="gpt-4o-realtime-preview-2024-10-01") as conn:
+        async with self.client.beta.realtime.connect(model="gpt-4o-realtime-preview") as conn:
             self.connection = conn
             self.connected.set()
 

Diff for: mypy.ini

@@ -44,7 +44,7 @@ cache_fine_grained = True
 # ```
 # Changing this codegen to make mypy happy would increase complexity
 # and would not be worth it.
-disable_error_code = func-returns-value
+disable_error_code = func-returns-value,overload-cannot-match
 
 # https://github.com/python/mypy/issues/12162
 [mypy.overrides]

Diff for: pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "openai"
-version = "1.59.6"
+version = "1.59.8"
 description = "The official Python library for the openai API"
 dynamic = ["readme"]
 license = "Apache-2.0"

Diff for: requirements-dev.lock

@@ -83,7 +83,7 @@ msal==1.31.0
     # via msal-extensions
 msal-extensions==1.2.0
     # via azure-identity
-mypy==1.13.0
+mypy==1.14.1
 mypy-extensions==1.0.0
     # via black
     # via mypy
@@ -124,7 +124,7 @@ pygments==2.18.0
     # via rich
 pyjwt==2.8.0
     # via msal
-pyright==1.1.390
+pyright==1.1.392.post0
 pytest==8.3.3
     # via pytest-asyncio
 pytest-asyncio==0.24.0

Diff for: src/openai/__init__.py

@@ -33,6 +33,7 @@
 )
 from ._base_client import DefaultHttpxClient, DefaultAsyncHttpxClient
 from ._utils._logs import setup_logging as _setup_logging
+from ._legacy_response import HttpxBinaryResponseContent as HttpxBinaryResponseContent
 
 __all__ = [
     "types",

Diff for: src/openai/_legacy_response.py

@@ -269,7 +269,9 @@ def _parse(self, *, to: type[_T] | None = None) -> R | _T:
         if origin == LegacyAPIResponse:
             raise RuntimeError("Unexpected state - cast_to is `APIResponse`")
 
-        if inspect.isclass(origin) and issubclass(origin, httpx.Response):
+        if inspect.isclass(
+            origin  # pyright: ignore[reportUnknownArgumentType]
+        ) and issubclass(origin, httpx.Response):
             # Because of the invariance of our ResponseT TypeVar, users can subclass httpx.Response
             # and pass that class to our request functions. We cannot change the variance to be either
             # covariant or contravariant as that makes our usage of ResponseT illegal. We could construct
@@ -279,7 +281,13 @@ def _parse(self, *, to: type[_T] | None = None) -> R | _T:
                 raise ValueError(f"Subclasses of httpx.Response cannot be passed to `cast_to`")
             return cast(R, response)
 
-        if inspect.isclass(origin) and not issubclass(origin, BaseModel) and issubclass(origin, pydantic.BaseModel):
+        if (
+            inspect.isclass(
+                origin  # pyright: ignore[reportUnknownArgumentType]
+            )
+            and not issubclass(origin, BaseModel)
+            and issubclass(origin, pydantic.BaseModel)
+        ):
             raise TypeError("Pydantic models must subclass our base model type, e.g. `from openai import BaseModel`")
 
         if (

Diff for: src/openai/_response.py

@@ -214,7 +214,13 @@ def _parse(self, *, to: type[_T] | None = None) -> R | _T:
                 raise ValueError(f"Subclasses of httpx.Response cannot be passed to `cast_to`")
             return cast(R, response)
 
-        if inspect.isclass(origin) and not issubclass(origin, BaseModel) and issubclass(origin, pydantic.BaseModel):
+        if (
+            inspect.isclass(
+                origin  # pyright: ignore[reportUnknownArgumentType]
+            )
+            and not issubclass(origin, BaseModel)
+            and issubclass(origin, pydantic.BaseModel)
+        ):
             raise TypeError("Pydantic models must subclass our base model type, e.g. `from openai import BaseModel`")
 
         if (

Diff for: src/openai/_version.py

@@ -1,4 +1,4 @@
 # File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details.
 
 __title__ = "openai"
-__version__ = "1.59.6"  # x-release-please-version
+__version__ = "1.59.8"  # x-release-please-version

Diff for: src/openai/lib/_parsing/_completions.py

@@ -157,7 +157,7 @@ def maybe_parse_content(
     response_format: type[ResponseFormatT] | ResponseFormatParam | NotGiven,
     message: ChatCompletionMessage | ParsedChatCompletionMessage[object],
 ) -> ResponseFormatT | None:
-    if has_rich_response_format(response_format) and message.content is not None and not message.refusal:
+    if has_rich_response_format(response_format) and message.content and not message.refusal:
         return _parse_content(response_format, message.content)
 
     return None

Diff for: src/openai/lib/_pydantic.py

@@ -108,6 +108,9 @@ def _ensure_strict_json_schema(
         # properties from the json schema take priority over the ones on the `$ref`
         json_schema.update({**resolved, **json_schema})
         json_schema.pop("$ref")
+        # Since the schema expanded from `$ref` might not have `additionalProperties: false` applied,
+        # we call `_ensure_strict_json_schema` again to fix the inlined schema and ensure it's valid.
+        return _ensure_strict_json_schema(json_schema, path=path, root=root)
 
     return json_schema
 

Diff for: src/openai/types/beta/__init__.py

@@ -43,3 +43,6 @@
 from .assistant_response_format_option_param import (
     AssistantResponseFormatOptionParam as AssistantResponseFormatOptionParam,
 )
+from .static_file_chunking_strategy_object_param import (
+    StaticFileChunkingStrategyObjectParam as StaticFileChunkingStrategyObjectParam,
+)

Diff for: src/openai/types/beta/file_chunking_strategy_param.py

@@ -6,8 +6,8 @@
 from typing_extensions import TypeAlias
 
 from .auto_file_chunking_strategy_param import AutoFileChunkingStrategyParam
-from .static_file_chunking_strategy_param import StaticFileChunkingStrategyParam
+from .static_file_chunking_strategy_object_param import StaticFileChunkingStrategyObjectParam
 
 __all__ = ["FileChunkingStrategyParam"]
 
-FileChunkingStrategyParam: TypeAlias = Union[AutoFileChunkingStrategyParam, StaticFileChunkingStrategyParam]
+FileChunkingStrategyParam: TypeAlias = Union[AutoFileChunkingStrategyParam, StaticFileChunkingStrategyObjectParam]

Diff for: src/openai/types/beta/static_file_chunking_strategy_object_param.py

@@ -0,0 +1,16 @@
+# File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details.
+
+from __future__ import annotations
+
+from typing_extensions import Literal, Required, TypedDict
+
+from .static_file_chunking_strategy_param import StaticFileChunkingStrategyParam
+
+__all__ = ["StaticFileChunkingStrategyObjectParam"]
+
+
+class StaticFileChunkingStrategyObjectParam(TypedDict, total=False):
+    static: Required[StaticFileChunkingStrategyParam]
+
+    type: Required[Literal["static"]]
+    """Always `static`."""