feat(api): update api shapes for usage and code interpreter

Author: stainless-app[bot]

.stats.yml

@@ -1,4 +1,4 @@
 configured_endpoints: 111
-openapi_spec_url: https://storage.googleapis.com/stainless-sdk-openapi-specs/openai%2Fopenai-f411a68f272b8be0ab0c266043da33228687b9b2d76896724e3cef797de9563d.yml
-openapi_spec_hash: 89bf866ea95ecfb3d76c8833237047d6
-config_hash: dc5515e257676a27cb1ace1784aa92b3
+openapi_spec_url: https://storage.googleapis.com/stainless-sdk-openapi-specs/openai%2Fopenai-ef4ecb19eb61e24c49d77fef769ee243e5279bc0bdbaee8d0f8dba4da8722559.yml
+openapi_spec_hash: 1b8a9767c9f04e6865b06c41948cdc24
+config_hash: fd2af1d5eff0995bb7dc02ac9a34851d

api.md

@@ -293,7 +293,7 @@ from openai.types.fine_tuning.checkpoints import (
 Methods:
 
 - <code title="post /fine_tuning/checkpoints/{fine_tuned_model_checkpoint}/permissions">client.fine_tuning.checkpoints.permissions.<a href="./src/openai/resources/fine_tuning/checkpoints/permissions.py">create</a>(fine_tuned_model_checkpoint, \*\*<a href="src/openai/types/fine_tuning/checkpoints/permission_create_params.py">params</a>) -> <a href="./src/openai/types/fine_tuning/checkpoints/permission_create_response.py">SyncPage[PermissionCreateResponse]</a></code>
-- <code title="get /fine_tuning/checkpoints/{fine_tuned_model_checkpoint}/permissions">client.fine_tuning.checkpoints.permissions.<a href="./src/openai/resources/fine_tuning/checkpoints/permissions.py">retrieve</a>(fine_tuned_model_checkpoint, \*\*<a href="src/openai/types/fine_tuning/checkpoints/permission_retrieve_params.py">params</a>) -> <a href="./src/openai/types/fine_tuning/checkpoints/permission_retrieve_response.py">SyncCursorPage[PermissionRetrieveResponse]</a></code>
+- <code title="get /fine_tuning/checkpoints/{fine_tuned_model_checkpoint}/permissions">client.fine_tuning.checkpoints.permissions.<a href="./src/openai/resources/fine_tuning/checkpoints/permissions.py">retrieve</a>(fine_tuned_model_checkpoint, \*\*<a href="src/openai/types/fine_tuning/checkpoints/permission_retrieve_params.py">params</a>) -> <a href="./src/openai/types/fine_tuning/checkpoints/permission_retrieve_response.py">PermissionRetrieveResponse</a></code>
 - <code title="delete /fine_tuning/checkpoints/{fine_tuned_model_checkpoint}/permissions/{permission_id}">client.fine_tuning.checkpoints.permissions.<a href="./src/openai/resources/fine_tuning/checkpoints/permissions.py">delete</a>(permission_id, \*, fine_tuned_model_checkpoint) -> <a href="./src/openai/types/fine_tuning/checkpoints/permission_delete_response.py">PermissionDeleteResponse</a></code>
 
 ## Alpha

src/openai/resources/audio/speech.py

@@ -56,6 +56,7 @@ def create(
         instructions: str | NotGiven = NOT_GIVEN,
         response_format: Literal["mp3", "opus", "aac", "flac", "wav", "pcm"] | NotGiven = NOT_GIVEN,
         speed: float | NotGiven = NOT_GIVEN,
+        stream_format: Literal["sse", "audio"] | NotGiven = NOT_GIVEN,
         # Use the following arguments if you need to pass additional parameters to the API that aren't available via kwargs.
         # The extra values given here take precedence over values defined on the client or passed to this method.
         extra_headers: Headers | None = None,
@@ -85,7 +86,10 @@ def create(
               `wav`, and `pcm`.
 
           speed: The speed of the generated audio. Select a value from `0.25` to `4.0`. `1.0` is
-              the default. Does not work with `gpt-4o-mini-tts`.
+              the default.
+
+          stream_format: The format to stream the audio in. Supported formats are `sse` and `audio`.
+              `sse` is not supported for `tts-1` or `tts-1-hd`.
 
           extra_headers: Send extra headers
 
@@ -106,6 +110,7 @@ def create(
                     "instructions": instructions,
                     "response_format": response_format,
                     "speed": speed,
+                    "stream_format": stream_format,
                 },
                 speech_create_params.SpeechCreateParams,
             ),
@@ -147,6 +152,7 @@ async def create(
         instructions: str | NotGiven = NOT_GIVEN,
         response_format: Literal["mp3", "opus", "aac", "flac", "wav", "pcm"] | NotGiven = NOT_GIVEN,
         speed: float | NotGiven = NOT_GIVEN,
+        stream_format: Literal["sse", "audio"] | NotGiven = NOT_GIVEN,
         # Use the following arguments if you need to pass additional parameters to the API that aren't available via kwargs.
         # The extra values given here take precedence over values defined on the client or passed to this method.
         extra_headers: Headers | None = None,
@@ -176,7 +182,10 @@ async def create(
               `wav`, and `pcm`.
 
           speed: The speed of the generated audio. Select a value from `0.25` to `4.0`. `1.0` is
-              the default. Does not work with `gpt-4o-mini-tts`.
+              the default.
+
+          stream_format: The format to stream the audio in. Supported formats are `sse` and `audio`.
+              `sse` is not supported for `tts-1` or `tts-1-hd`.
 
           extra_headers: Send extra headers
 
@@ -197,6 +206,7 @@ async def create(
                     "instructions": instructions,
                     "response_format": response_format,
                     "speed": speed,
+                    "stream_format": stream_format,
                 },
                 speech_create_params.SpeechCreateParams,
             ),

src/openai/resources/fine_tuning/checkpoints/permissions.py

@@ -9,11 +9,11 @@
 
 from .... import _legacy_response
 from ...._types import NOT_GIVEN, Body, Query, Headers, NotGiven
-from ...._utils import maybe_transform
+from ...._utils import maybe_transform, async_maybe_transform
 from ...._compat import cached_property
 from ...._resource import SyncAPIResource, AsyncAPIResource
 from ...._response import to_streamed_response_wrapper, async_to_streamed_response_wrapper
-from ....pagination import SyncPage, AsyncPage, SyncCursorPage, AsyncCursorPage
+from ....pagination import SyncPage, AsyncPage
 from ...._base_client import AsyncPaginator, make_request_options
 from ....types.fine_tuning.checkpoints import permission_create_params, permission_retrieve_params
 from ....types.fine_tuning.checkpoints.permission_create_response import PermissionCreateResponse
@@ -101,7 +101,7 @@ def retrieve(
         extra_query: Query | None = None,
         extra_body: Body | None = None,
         timeout: float | httpx.Timeout | None | NotGiven = NOT_GIVEN,
-    ) -> SyncCursorPage[PermissionRetrieveResponse]:
+    ) -> PermissionRetrieveResponse:
         """
         **NOTE:** This endpoint requires an [admin API key](../admin-api-keys).
 
@@ -129,9 +129,8 @@ def retrieve(
             raise ValueError(
                 f"Expected a non-empty value for `fine_tuned_model_checkpoint` but received {fine_tuned_model_checkpoint!r}"
             )
-        return self._get_api_list(
+        return self._get(
             f"/fine_tuning/checkpoints/{fine_tuned_model_checkpoint}/permissions",
-            page=SyncCursorPage[PermissionRetrieveResponse],
             options=make_request_options(
                 extra_headers=extra_headers,
                 extra_query=extra_query,
@@ -147,7 +146,7 @@ def retrieve(
                     permission_retrieve_params.PermissionRetrieveParams,
                 ),
             ),
-            model=PermissionRetrieveResponse,
+            cast_to=PermissionRetrieveResponse,
         )
 
     def delete(
@@ -256,7 +255,7 @@ def create(
             method="post",
         )
 
-    def retrieve(
+    async def retrieve(
         self,
         fine_tuned_model_checkpoint: str,
         *,
@@ -270,7 +269,7 @@ def retrieve(
         extra_query: Query | None = None,
         extra_body: Body | None = None,
         timeout: float | httpx.Timeout | None | NotGiven = NOT_GIVEN,
-    ) -> AsyncPaginator[PermissionRetrieveResponse, AsyncCursorPage[PermissionRetrieveResponse]]:
+    ) -> PermissionRetrieveResponse:
         """
         **NOTE:** This endpoint requires an [admin API key](../admin-api-keys).
 
@@ -298,15 +297,14 @@ def retrieve(
             raise ValueError(
                 f"Expected a non-empty value for `fine_tuned_model_checkpoint` but received {fine_tuned_model_checkpoint!r}"
             )
-        return self._get_api_list(
+        return await self._get(
             f"/fine_tuning/checkpoints/{fine_tuned_model_checkpoint}/permissions",
-            page=AsyncCursorPage[PermissionRetrieveResponse],
             options=make_request_options(
                 extra_headers=extra_headers,
                 extra_query=extra_query,
                 extra_body=extra_body,
                 timeout=timeout,
-                query=maybe_transform(
+                query=await async_maybe_transform(
                     {
                         "after": after,
                         "limit": limit,
@@ -316,7 +314,7 @@ def retrieve(
                     permission_retrieve_params.PermissionRetrieveParams,
                 ),
             ),
-            model=PermissionRetrieveResponse,
+            cast_to=PermissionRetrieveResponse,
         )
 
     async def delete(

src/openai/types/audio/speech_create_params.py

@@ -48,6 +48,12 @@ class SpeechCreateParams(TypedDict, total=False):
     speed: float
     """The speed of the generated audio.
 
-    Select a value from `0.25` to `4.0`. `1.0` is the default. Does not work with
-    `gpt-4o-mini-tts`.
+    Select a value from `0.25` to `4.0`. `1.0` is the default.
+    """
+
+    stream_format: Literal["sse", "audio"]
+    """The format to stream the audio in.
+
+    Supported formats are `sse` and `audio`. `sse` is not supported for `tts-1` or
+    `tts-1-hd`.
     """

src/openai/types/audio/transcription.py

@@ -1,10 +1,12 @@
 # File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details.
 
-from typing import List, Optional
+from typing import List, Union, Optional
+from typing_extensions import Literal, Annotated, TypeAlias
 
+from ..._utils import PropertyInfo
 from ..._models import BaseModel
 
-__all__ = ["Transcription", "Logprob"]
+__all__ = ["Transcription", "Logprob", "Usage", "UsageTokens", "UsageTokensInputTokenDetails", "UsageDuration"]
 
 
 class Logprob(BaseModel):
@@ -18,6 +20,42 @@ class Logprob(BaseModel):
     """The log probability of the token."""
 
 
+class UsageTokensInputTokenDetails(BaseModel):
+    audio_tokens: Optional[int] = None
+    """Number of audio tokens billed for this request."""
+
+    text_tokens: Optional[int] = None
+    """Number of text tokens billed for this request."""
+
+
+class UsageTokens(BaseModel):
+    input_tokens: int
+    """Number of input tokens billed for this request."""
+
+    output_tokens: int
+    """Number of output tokens generated."""
+
+    total_tokens: int
+    """Total number of tokens used (input + output)."""
+
+    type: Literal["tokens"]
+    """The type of the usage object. Always `tokens` for this variant."""
+
+    input_token_details: Optional[UsageTokensInputTokenDetails] = None
+    """Details about the input tokens billed for this request."""
+
+
+class UsageDuration(BaseModel):
+    duration: float
+    """Duration of the input audio in seconds."""
+
+    type: Literal["duration"]
+    """The type of the usage object. Always `duration` for this variant."""
+
+
+Usage: TypeAlias = Annotated[Union[UsageTokens, UsageDuration], PropertyInfo(discriminator="type")]
+
+
 class Transcription(BaseModel):
     text: str
     """The transcribed text."""
@@ -28,3 +66,6 @@ class Transcription(BaseModel):
     Only returned with the models `gpt-4o-transcribe` and `gpt-4o-mini-transcribe`
     if `logprobs` is added to the `include` array.
     """
+
+    usage: Optional[Usage] = None
+    """Token usage statistics for the request."""

src/openai/types/audio/transcription_text_done_event.py

@@ -5,7 +5,7 @@
 
 from ..._models import BaseModel
 
-__all__ = ["TranscriptionTextDoneEvent", "Logprob"]
+__all__ = ["TranscriptionTextDoneEvent", "Logprob", "Usage", "UsageInputTokenDetails"]
 
 
 class Logprob(BaseModel):
@@ -19,6 +19,31 @@ class Logprob(BaseModel):
     """The log probability of the token."""
 
 
+class UsageInputTokenDetails(BaseModel):
+    audio_tokens: Optional[int] = None
+    """Number of audio tokens billed for this request."""
+
+    text_tokens: Optional[int] = None
+    """Number of text tokens billed for this request."""
+
+
+class Usage(BaseModel):
+    input_tokens: int
+    """Number of input tokens billed for this request."""
+
+    output_tokens: int
+    """Number of output tokens generated."""
+
+    total_tokens: int
+    """Total number of tokens used (input + output)."""
+
+    type: Literal["tokens"]
+    """The type of the usage object. Always `tokens` for this variant."""
+
+    input_token_details: Optional[UsageInputTokenDetails] = None
+    """Details about the input tokens billed for this request."""
+
+
 class TranscriptionTextDoneEvent(BaseModel):
     text: str
     """The text that was transcribed."""
@@ -33,3 +58,6 @@ class TranscriptionTextDoneEvent(BaseModel):
     [create a transcription](https://platform.openai.com/docs/api-reference/audio/create-transcription)
     with the `include[]` parameter set to `logprobs`.
     """
+
+    usage: Optional[Usage] = None
+    """Usage statistics for models billed by token usage."""

src/openai/types/audio/transcription_verbose.py

@@ -1,12 +1,21 @@
 # File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details.
 
 from typing import List, Optional
+from typing_extensions import Literal
 
 from ..._models import BaseModel
 from .transcription_word import TranscriptionWord
 from .transcription_segment import TranscriptionSegment
 
-__all__ = ["TranscriptionVerbose"]
+__all__ = ["TranscriptionVerbose", "Usage"]
+
+
+class Usage(BaseModel):
+    duration: float
+    """Duration of the input audio in seconds."""
+
+    type: Literal["duration"]
+    """The type of the usage object. Always `duration` for this variant."""
 
 
 class TranscriptionVerbose(BaseModel):
@@ -22,5 +31,8 @@ class TranscriptionVerbose(BaseModel):
     segments: Optional[List[TranscriptionSegment]] = None
     """Segments of the transcribed text and their corresponding details."""
 
+    usage: Optional[Usage] = None
+    """Usage statistics for models billed by audio input duration."""
+
     words: Optional[List[TranscriptionWord]] = None
     """Extracted words and their corresponding timestamps."""

src/openai/types/beta/realtime/session_create_params.py

@@ -3,12 +3,12 @@
 from __future__ import annotations
 
 from typing import List, Union, Iterable
-from typing_extensions import Literal, TypeAlias, TypedDict
+from typing_extensions import Literal, Required, TypeAlias, TypedDict
 
 __all__ = [
     "SessionCreateParams",
     "ClientSecret",
-    "ClientSecretExpiresAt",
+    "ClientSecretExpiresAfter",
     "InputAudioNoiseReduction",
     "InputAudioTranscription",
     "Tool",
@@ -156,8 +156,8 @@ class SessionCreateParams(TypedDict, total=False):
     """
 
 
-class ClientSecretExpiresAt(TypedDict, total=False):
-    anchor: Literal["created_at"]
+class ClientSecretExpiresAfter(TypedDict, total=False):
+    anchor: Required[Literal["created_at"]]
     """The anchor point for the ephemeral token expiration.
 
     Only `created_at` is currently supported.
@@ -171,7 +171,7 @@ class ClientSecretExpiresAt(TypedDict, total=False):
 
 
 class ClientSecret(TypedDict, total=False):
-    expires_at: ClientSecretExpiresAt
+    expires_after: ClientSecretExpiresAfter
     """Configuration for the ephemeral token expiration."""
 
 

src/openai/types/beta/realtime/session_create_response.py

@@ -33,10 +33,7 @@ class ClientSecret(BaseModel):
 
 class InputAudioTranscription(BaseModel):
     model: Optional[str] = None
-    """
-    The model to use for transcription, `whisper-1` is the only currently supported
-    model.
-    """
+    """The model to use for transcription."""
 
 
 class Tool(BaseModel):
@@ -116,8 +113,8 @@ class SessionCreateResponse(BaseModel):
     Configuration for input audio transcription, defaults to off and can be set to
     `null` to turn off once on. Input audio transcription is not native to the
     model, since the model consumes audio directly. Transcription runs
-    asynchronously through Whisper and should be treated as rough guidance rather
-    than the representation understood by the model.
+    asynchronously and should be treated as rough guidance rather than the
+    representation understood by the model.
     """
 
     instructions: Optional[str] = None