release: 2.3.0

.release-please-manifest.json

@@ -1,3 +1,3 @@
 {
-  ".": "2.2.0"
+  ".": "2.3.0"
 }

.stats.yml

@@ -1,4 +1,4 @@
 configured_endpoints: 136
-openapi_spec_url: https://storage.googleapis.com/stainless-sdk-openapi-specs/openai%2Fopenai-d64cf80d2ebddf175c5578f68226a3d5bbd3f7fd8d62ccac2205f3fc05a355ee.yml
-openapi_spec_hash: d51e0d60d0c536f210b597a211bc5af0
-config_hash: e7c42016df9c6bd7bd6ff15101b9bc9b
+openapi_spec_url: https://storage.googleapis.com/stainless-sdk-openapi-specs/openai%2Fopenai-e66e85fb7f72477256dca1acb6b23396989d381c5c1b318de564195436bcb93f.yml
+openapi_spec_hash: 0a4bbb5aa0ae532a072bd6b3854e70b1
+config_hash: 89bf7bb3a1f9439ffc6ea0e7dc57ba9b

CHANGELOG.md

@@ -1,5 +1,18 @@
 # Changelog
 
+## 2.3.0 (2025-10-10)
+
+Full Changelog: [v2.2.0...v2.3.0](https://github.com/openai/openai-python/compare/v2.2.0...v2.3.0)
+
+### Features
+
+* **api:** comparison filter in/not in ([aa49f62](https://github.com/openai/openai-python/commit/aa49f626a6ea9d77ad008badfb3741e16232d62f))
+
+
+### Chores
+
+* **package:** bump jiter to >=0.10.0 to support Python 3.14 ([#2618](https://github.com/openai/openai-python/issues/2618)) ([aa445ca](https://github.com/openai/openai-python/commit/aa445cab5c93c6908697fe98e73e16963330b141))
+
 ## 2.2.0 (2025-10-06)
 
 Full Changelog: [v2.1.0...v2.2.0](https://github.com/openai/openai-python/compare/v2.1.0...v2.2.0)

pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "openai"
-version = "2.2.0"
+version = "2.3.0"
 description = "The official Python library for the openai API"
 dynamic = ["readme"]
 license = "Apache-2.0"
@@ -15,7 +15,7 @@ dependencies = [
     "distro>=1.7.0, <2",
     "sniffio",
     "tqdm > 4",
-    "jiter>=0.4.0, <1",
+    "jiter>=0.10.0, <1",
 ]
 requires-python = ">= 3.8"
 classifiers = [

requirements-dev.lock

@@ -91,7 +91,7 @@ importlib-metadata==7.0.0
 iniconfig==2.0.0
     # via pytest
 inline-snapshot==0.28.0
-jiter==0.5.0
+jiter==0.11.0
     # via openai
 markdown-it-py==3.0.0
     # via rich

requirements.lock

@@ -51,7 +51,7 @@ idna==3.4
     # via anyio
     # via httpx
     # via yarl
-jiter==0.6.1
+jiter==0.11.0
     # via openai
 multidict==6.5.0
     # via aiohttp

src/openai/_version.py

@@ -1,4 +1,4 @@
 # File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details.
 
 __title__ = "openai"
-__version__ = "2.2.0"  # x-release-please-version
+__version__ = "2.3.0"  # x-release-please-version

src/openai/resources/beta/assistants.py

@@ -102,6 +102,9 @@ def create(
               effort can result in faster responses and fewer tokens used on reasoning in a
               response.
 
+              Note: The `gpt-5-pro` model defaults to (and only supports) `high` reasoning
+              effort.
+
           response_format: Specifies the format that the model must output. Compatible with
               [GPT-4o](https://platform.openai.com/docs/models#gpt-4o),
               [GPT-4 Turbo](https://platform.openai.com/docs/models#gpt-4-turbo-and-gpt-4),
@@ -309,6 +312,9 @@ def update(
               effort can result in faster responses and fewer tokens used on reasoning in a
               response.
 
+              Note: The `gpt-5-pro` model defaults to (and only supports) `high` reasoning
+              effort.
+
           response_format: Specifies the format that the model must output. Compatible with
               [GPT-4o](https://platform.openai.com/docs/models#gpt-4o),
               [GPT-4 Turbo](https://platform.openai.com/docs/models#gpt-4-turbo-and-gpt-4),
@@ -555,6 +561,9 @@ async def create(
               effort can result in faster responses and fewer tokens used on reasoning in a
               response.
 
+              Note: The `gpt-5-pro` model defaults to (and only supports) `high` reasoning
+              effort.
+
           response_format: Specifies the format that the model must output. Compatible with
               [GPT-4o](https://platform.openai.com/docs/models#gpt-4o),
               [GPT-4 Turbo](https://platform.openai.com/docs/models#gpt-4-turbo-and-gpt-4),
@@ -762,6 +771,9 @@ async def update(
               effort can result in faster responses and fewer tokens used on reasoning in a
               response.
 
+              Note: The `gpt-5-pro` model defaults to (and only supports) `high` reasoning
+              effort.
+
           response_format: Specifies the format that the model must output. Compatible with
               [GPT-4o](https://platform.openai.com/docs/models#gpt-4o),
               [GPT-4 Turbo](https://platform.openai.com/docs/models#gpt-4-turbo-and-gpt-4),

src/openai/resources/beta/threads/runs/runs.py

@@ -173,6 +173,9 @@ def create(
               effort can result in faster responses and fewer tokens used on reasoning in a
               response.
 
+              Note: The `gpt-5-pro` model defaults to (and only supports) `high` reasoning
+              effort.
+
           response_format: Specifies the format that the model must output. Compatible with
               [GPT-4o](https://platform.openai.com/docs/models#gpt-4o),
               [GPT-4 Turbo](https://platform.openai.com/docs/models#gpt-4-turbo-and-gpt-4),
@@ -327,6 +330,9 @@ def create(
               effort can result in faster responses and fewer tokens used on reasoning in a
               response.
 
+              Note: The `gpt-5-pro` model defaults to (and only supports) `high` reasoning
+              effort.
+
           response_format: Specifies the format that the model must output. Compatible with
               [GPT-4o](https://platform.openai.com/docs/models#gpt-4o),
               [GPT-4 Turbo](https://platform.openai.com/docs/models#gpt-4-turbo-and-gpt-4),
@@ -477,6 +483,9 @@ def create(
               effort can result in faster responses and fewer tokens used on reasoning in a
               response.
 
+              Note: The `gpt-5-pro` model defaults to (and only supports) `high` reasoning
+              effort.
+
           response_format: Specifies the format that the model must output. Compatible with
               [GPT-4o](https://platform.openai.com/docs/models#gpt-4o),
               [GPT-4 Turbo](https://platform.openai.com/docs/models#gpt-4-turbo-and-gpt-4),
@@ -1603,6 +1612,9 @@ async def create(
               effort can result in faster responses and fewer tokens used on reasoning in a
               response.
 
+              Note: The `gpt-5-pro` model defaults to (and only supports) `high` reasoning
+              effort.
+
           response_format: Specifies the format that the model must output. Compatible with
               [GPT-4o](https://platform.openai.com/docs/models#gpt-4o),
               [GPT-4 Turbo](https://platform.openai.com/docs/models#gpt-4-turbo-and-gpt-4),
@@ -1757,6 +1769,9 @@ async def create(
               effort can result in faster responses and fewer tokens used on reasoning in a
               response.
 
+              Note: The `gpt-5-pro` model defaults to (and only supports) `high` reasoning
+              effort.
+
           response_format: Specifies the format that the model must output. Compatible with
               [GPT-4o](https://platform.openai.com/docs/models#gpt-4o),
               [GPT-4 Turbo](https://platform.openai.com/docs/models#gpt-4-turbo-and-gpt-4),
@@ -1907,6 +1922,9 @@ async def create(
               effort can result in faster responses and fewer tokens used on reasoning in a
               response.
 
+              Note: The `gpt-5-pro` model defaults to (and only supports) `high` reasoning
+              effort.
+
           response_format: Specifies the format that the model must output. Compatible with
               [GPT-4o](https://platform.openai.com/docs/models#gpt-4o),
               [GPT-4 Turbo](https://platform.openai.com/docs/models#gpt-4-turbo-and-gpt-4),

src/openai/resources/chat/completions/completions.py

@@ -407,6 +407,9 @@ def create(
               effort can result in faster responses and fewer tokens used on reasoning in a
               response.
 
+              Note: The `gpt-5-pro` model defaults to (and only supports) `high` reasoning
+              effort.
+
           response_format: An object specifying the format that the model must output.
 
               Setting to `{ "type": "json_schema", "json_schema": {...} }` enables Structured
@@ -704,6 +707,9 @@ def create(
               effort can result in faster responses and fewer tokens used on reasoning in a
               response.
 
+              Note: The `gpt-5-pro` model defaults to (and only supports) `high` reasoning
+              effort.
+
           response_format: An object specifying the format that the model must output.
 
               Setting to `{ "type": "json_schema", "json_schema": {...} }` enables Structured
@@ -992,6 +998,9 @@ def create(
               effort can result in faster responses and fewer tokens used on reasoning in a
               response.
 
+              Note: The `gpt-5-pro` model defaults to (and only supports) `high` reasoning
+              effort.
+
           response_format: An object specifying the format that the model must output.
 
               Setting to `{ "type": "json_schema", "json_schema": {...} }` enables Structured
@@ -1845,6 +1854,9 @@ async def create(
               effort can result in faster responses and fewer tokens used on reasoning in a
               response.
 
+              Note: The `gpt-5-pro` model defaults to (and only supports) `high` reasoning
+              effort.
+
           response_format: An object specifying the format that the model must output.
 
               Setting to `{ "type": "json_schema", "json_schema": {...} }` enables Structured
@@ -2142,6 +2154,9 @@ async def create(
               effort can result in faster responses and fewer tokens used on reasoning in a
               response.
 
+              Note: The `gpt-5-pro` model defaults to (and only supports) `high` reasoning
+              effort.
+
           response_format: An object specifying the format that the model must output.
 
               Setting to `{ "type": "json_schema", "json_schema": {...} }` enables Structured
@@ -2430,6 +2445,9 @@ async def create(
               effort can result in faster responses and fewer tokens used on reasoning in a
               response.
 
+              Note: The `gpt-5-pro` model defaults to (and only supports) `high` reasoning
+              effort.
+
           response_format: An object specifying the format that the model must output.
 
               Setting to `{ "type": "json_schema", "json_schema": {...} }` enables Structured

src/openai/resources/files.py

@@ -236,7 +236,7 @@ def delete(
         timeout: float | httpx.Timeout | None | NotGiven = not_given,
     ) -> FileDeleted:
         """
-        Delete a file.
+        Delete a file and remove it from all vector stores.
 
         Args:
           extra_headers: Send extra headers
@@ -553,7 +553,7 @@ async def delete(
         timeout: float | httpx.Timeout | None | NotGiven = not_given,
     ) -> FileDeleted:
         """
-        Delete a file.
+        Delete a file and remove it from all vector stores.
 
         Args:
           extra_headers: Send extra headers

src/openai/types/beta/assistant_create_params.py

@@ -65,6 +65,9 @@ class AssistantCreateParams(TypedDict, total=False):
     supported values are `minimal`, `low`, `medium`, and `high`. Reducing reasoning
     effort can result in faster responses and fewer tokens used on reasoning in a
     response.
+
+    Note: The `gpt-5-pro` model defaults to (and only supports) `high` reasoning
+    effort.
     """
 
     response_format: Optional[AssistantResponseFormatOptionParam]

src/openai/types/beta/assistant_update_params.py

@@ -100,6 +100,9 @@ class AssistantUpdateParams(TypedDict, total=False):
     supported values are `minimal`, `low`, `medium`, and `high`. Reducing reasoning
     effort can result in faster responses and fewer tokens used on reasoning in a
     response.
+
+    Note: The `gpt-5-pro` model defaults to (and only supports) `high` reasoning
+    effort.
     """
 
     response_format: Optional[AssistantResponseFormatOptionParam]

src/openai/types/beta/threads/run_create_params.py

@@ -114,6 +114,9 @@ class RunCreateParamsBase(TypedDict, total=False):
     supported values are `minimal`, `low`, `medium`, and `high`. Reducing reasoning
     effort can result in faster responses and fewer tokens used on reasoning in a
     response.
+
+    Note: The `gpt-5-pro` model defaults to (and only supports) `high` reasoning
+    effort.
     """
 
     response_format: Optional[AssistantResponseFormatOptionParam]

src/openai/types/chat/completion_create_params.py

@@ -192,6 +192,9 @@ class CompletionCreateParamsBase(TypedDict, total=False):
     supported values are `minimal`, `low`, `medium`, and `high`. Reducing reasoning
     effort can result in faster responses and fewer tokens used on reasoning in a
     response.
+
+    Note: The `gpt-5-pro` model defaults to (and only supports) `high` reasoning
+    effort.
     """
 
     response_format: ResponseFormat

src/openai/types/evals/create_eval_completions_run_data_source.py

@@ -175,6 +175,9 @@ class SamplingParams(BaseModel):
     supported values are `minimal`, `low`, `medium`, and `high`. Reducing reasoning
     effort can result in faster responses and fewer tokens used on reasoning in a
     response.
+
+    Note: The `gpt-5-pro` model defaults to (and only supports) `high` reasoning
+    effort.
     """
 
     response_format: Optional[SamplingParamsResponseFormat] = None

src/openai/types/evals/create_eval_completions_run_data_source_param.py

@@ -171,6 +171,9 @@ class SamplingParams(TypedDict, total=False):
     supported values are `minimal`, `low`, `medium`, and `high`. Reducing reasoning
     effort can result in faster responses and fewer tokens used on reasoning in a
     response.
+
+    Note: The `gpt-5-pro` model defaults to (and only supports) `high` reasoning
+    effort.
     """
 
     response_format: SamplingParamsResponseFormat

src/openai/types/evals/run_cancel_response.py

@@ -106,6 +106,9 @@ class DataSourceResponsesSourceResponses(BaseModel):
     supported values are `minimal`, `low`, `medium`, and `high`. Reducing reasoning
     effort can result in faster responses and fewer tokens used on reasoning in a
     response.
+
+    Note: The `gpt-5-pro` model defaults to (and only supports) `high` reasoning
+    effort.
     """
 
     temperature: Optional[float] = None
@@ -241,6 +244,9 @@ class DataSourceResponsesSamplingParams(BaseModel):
     supported values are `minimal`, `low`, `medium`, and `high`. Reducing reasoning
     effort can result in faster responses and fewer tokens used on reasoning in a
     response.
+
+    Note: The `gpt-5-pro` model defaults to (and only supports) `high` reasoning
+    effort.
     """
 
     seed: Optional[int] = None

src/openai/types/evals/run_create_params.py

@@ -119,6 +119,9 @@ class DataSourceCreateEvalResponsesRunDataSourceSourceResponses(TypedDict, total
     supported values are `minimal`, `low`, `medium`, and `high`. Reducing reasoning
     effort can result in faster responses and fewer tokens used on reasoning in a
     response.
+
+    Note: The `gpt-5-pro` model defaults to (and only supports) `high` reasoning
+    effort.
     """
 
     temperature: Optional[float]
@@ -259,6 +262,9 @@ class DataSourceCreateEvalResponsesRunDataSourceSamplingParams(TypedDict, total=
     supported values are `minimal`, `low`, `medium`, and `high`. Reducing reasoning
     effort can result in faster responses and fewer tokens used on reasoning in a
     response.
+
+    Note: The `gpt-5-pro` model defaults to (and only supports) `high` reasoning
+    effort.
     """
 
     seed: int

src/openai/types/evals/run_create_response.py

@@ -106,6 +106,9 @@ class DataSourceResponsesSourceResponses(BaseModel):
     supported values are `minimal`, `low`, `medium`, and `high`. Reducing reasoning
     effort can result in faster responses and fewer tokens used on reasoning in a
     response.
+
+    Note: The `gpt-5-pro` model defaults to (and only supports) `high` reasoning
+    effort.
     """
 
     temperature: Optional[float] = None
@@ -241,6 +244,9 @@ class DataSourceResponsesSamplingParams(BaseModel):
     supported values are `minimal`, `low`, `medium`, and `high`. Reducing reasoning
     effort can result in faster responses and fewer tokens used on reasoning in a
     response.
+
+    Note: The `gpt-5-pro` model defaults to (and only supports) `high` reasoning
+    effort.
     """
 
     seed: Optional[int] = None

src/openai/types/evals/run_list_response.py

@@ -106,6 +106,9 @@ class DataSourceResponsesSourceResponses(BaseModel):
     supported values are `minimal`, `low`, `medium`, and `high`. Reducing reasoning
     effort can result in faster responses and fewer tokens used on reasoning in a
     response.
+
+    Note: The `gpt-5-pro` model defaults to (and only supports) `high` reasoning
+    effort.
     """
 
     temperature: Optional[float] = None
@@ -241,6 +244,9 @@ class DataSourceResponsesSamplingParams(BaseModel):
     supported values are `minimal`, `low`, `medium`, and `high`. Reducing reasoning
     effort can result in faster responses and fewer tokens used on reasoning in a
     response.
+
+    Note: The `gpt-5-pro` model defaults to (and only supports) `high` reasoning
+    effort.
     """
 
     seed: Optional[int] = None