c63a5cfe-b235-4fbe-8bbb-82a9e02a482a-python 0.1.0a8__py3-none-any.whl → 0.1.0a9__py3-none-any.whl

{c63a5cfe_b235_4fbe_8bbb_82a9e02a482a_python-0.1.0a8.dist-info → c63a5cfe_b235_4fbe_8bbb_82a9e02a482a_python-0.1.0a9.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: c63a5cfe-b235-4fbe-8bbb-82a9e02a482a-python
-Version: 0.1.0a8
+Version: 0.1.0a9
 Summary: The official Python library for GradientAI
 Project-URL: Homepage, https://github.com/digitalocean/gradientai-python
 Project-URL: Repository, https://github.com/digitalocean/gradientai-python

{c63a5cfe_b235_4fbe_8bbb_82a9e02a482a_python-0.1.0a8.dist-info → c63a5cfe_b235_4fbe_8bbb_82a9e02a482a_python-0.1.0a9.dist-info}/RECORD

@@ -11,7 +11,7 @@ gradientai/_resource.py,sha256=4NZbH2h8dQ-t-DQPida4VANJ_oZJNA7qxV84mwJT8oM,1124
 gradientai/_response.py,sha256=RhlDdupxTcKNyDDj045MZD3-a_lsEc3yjiOzxWg0cDc,28842
 gradientai/_streaming.py,sha256=3KH-GBmqhoS1KAOhecADOsbW9WuzhIi8wSdmrEj5PPA,11404
 gradientai/_types.py,sha256=22gBoIuoGJ1R6l5nPwquWCRzJodKhO-3e7k22-h37JQ,6201
-gradientai/_version.py,sha256=vl2rc2vAkezojVTxFct2GJxzipzCGzHOxgS9Ld5ASmM,170
+gradientai/_version.py,sha256=aATDhXxfQvFSrVaXTkz4f895b3kBEp1bAJcYPj23INc,170
 gradientai/py.typed,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 gradientai/_utils/__init__.py,sha256=PNZ_QJuzZEgyYXqkO1HVhGkj5IU9bglVUcw7H-Knjzw,2062
 gradientai/_utils/_logs.py,sha256=1QuZcxzSKHcqNFbPoz_pvfLD5eDfOMBzGMyanAm_2gw,787
@@ -39,7 +39,7 @@ gradientai/resources/agents/routes.py,sha256=pDoK5hUgluhz5awJ9FR_krF8DIbkRzLX__M
 gradientai/resources/agents/versions.py,sha256=hmHDaawFkpqoUEYI4Vi5jckH3mamkPKupXO3cZccyKE,11546
 gradientai/resources/agents/chat/__init__.py,sha256=BVAfz9TM3DT5W9f_mt0P9YRxL_MsUxKCWAH6u1iogmA,1041
 gradientai/resources/agents/chat/chat.py,sha256=nt97777qa-xM71JQBKDFG_x7fUDJRvy6rF5LoBiMOKE,3698
-gradientai/resources/agents/chat/completions.py,sha256=jq62v8gN3hJ7POxBCHMcUJUi6Zj4IgvNGlr7D36W1M8,46188
+gradientai/resources/agents/chat/completions.py,sha256=uYlI68LprqXL-W1hVRjd0tb3jutXYosKKMOEecuzpV8,46239
 gradientai/resources/agents/evaluation_metrics/__init__.py,sha256=qUCsT_vI2TrZrUBPe8h-VMNBE4ytcoe0RXshDctV0g0,1198
 gradientai/resources/agents/evaluation_metrics/evaluation_metrics.py,sha256=BueqcWT0lqbElo-lgntkr1s7J0-qi2KqEZV2qzc7dcI,7089
 gradientai/resources/agents/evaluation_metrics/workspaces/__init__.py,sha256=Zf3wa7WSVOKyG1x9Fppny38_ewQCbdghxbptqGFVfOM,1054
@@ -47,7 +47,7 @@ gradientai/resources/agents/evaluation_metrics/workspaces/agents.py,sha256=AVgDN
 gradientai/resources/agents/evaluation_metrics/workspaces/workspaces.py,sha256=iSQS3IW5xjOX5EsyJraoS5enwzrBARUhm6Xl1EgiICE,27079
 gradientai/resources/chat/__init__.py,sha256=BVAfz9TM3DT5W9f_mt0P9YRxL_MsUxKCWAH6u1iogmA,1041
 gradientai/resources/chat/chat.py,sha256=ANOEzzDATXj-onkkdtvVad9O-Dfwe7Uza7yfA_3bU4U,3696
-gradientai/resources/chat/completions.py,sha256=KWtB8ro2ClciI4NhbXzgXL-bIE7fcInDdGR7EElA5nU,19108
+gradientai/resources/chat/completions.py,sha256=u_TvPvwNU79e3I5-v1P384LE5ysVz9gjJpYOKxnkl7E,47059
 gradientai/resources/inference/__init__.py,sha256=5Yk9bdOpUJPTero0_CvA-GJvcU5_gVlN1jf5r2GGGPY,1055
 gradientai/resources/inference/api_keys.py,sha256=xgWTFTp8IVRkw2nvXnfjBZPOTJyLUDIKlE7fSmb1y2I,22021
 gradientai/resources/inference/inference.py,sha256=jBimuBx3kKsuwC3pgy-PPrWg1TryO_a108CC_xLS2-Y,3771
@@ -150,8 +150,8 @@ gradientai/types/agents/version_list_params.py,sha256=0_3DhUbFDRyjUnn2G4saFOvuot
 gradientai/types/agents/version_list_response.py,sha256=Y0Y8CSPUPVHKRA3zTRRfQ8gC2aFyBacSCaGz-D5k8nk,2725
 gradientai/types/agents/version_update_params.py,sha256=j9tOda5wXmSOHsmcxQONo2mM-hEtrYi5-19HfGU_XnI,379
 gradientai/types/agents/version_update_response.py,sha256=nspPIkxQskT82tcW0JyG7bBVlXq_KU6CZzodTd9jfkQ,709
-gradientai/types/agents/chat/__init__.py,sha256=c-PmEwuvWZQ4CRBTs9gzbKAq2sxL7V7JlVxddeoaGl0,381
-gradientai/types/agents/chat/chat_completion_chunk.py,sha256=1K-F0JdUmQ_4idDk3oUGUB_mhxAxCzjq0C8hMhGtDuY,3048
+gradientai/types/agents/chat/__init__.py,sha256=VGhGOtQE4pcOIHKsOeLDYDpixc_5ExOfIJcR0q6AqvY,397
+gradientai/types/agents/chat/agent_chat_completion_chunk.py,sha256=bktrs-Ao6uta3k4PVzBrreMKbArztBcsHIEa8uzIK9s,3058
 gradientai/types/agents/chat/completion_create_params.py,sha256=ADEJ0N3MMsouT9AqBLE1-rho4FVVhlp9U9E-buqIAYs,7165
 gradientai/types/agents/chat/completion_create_response.py,sha256=9uKS3memEoV0_Xd1CZwI0jQGsQyfVkhXRPGlRO3rUIc,2415
 gradientai/types/agents/evaluation_metrics/__init__.py,sha256=XWH_utxMx-JwArRpr-rHQfmoxQRGK6GciKOllbkqg40,894
@@ -168,8 +168,9 @@ gradientai/types/agents/evaluation_metrics/workspaces/agent_list_params.py,sha25
 gradientai/types/agents/evaluation_metrics/workspaces/agent_list_response.py,sha256=W4O5v1LoWh2hQJTpUm5SfQCYcv6Q9Yz1Id5Pm1sPQNA,503
 gradientai/types/agents/evaluation_metrics/workspaces/agent_move_params.py,sha256=4INiLEvgT9UDqFbrGwp3nuWOzFhwv7sX_YCr1Um1RaQ,422
 gradientai/types/agents/evaluation_metrics/workspaces/agent_move_response.py,sha256=j2uoTmFwnLNU7kGI5LZdPpMUxWmUk-HdQIVVDOwSy4Y,350
-gradientai/types/chat/__init__.py,sha256=A5VCUPqJZydjjOqEXC01GXmcDkKM3bq6zuCu9lmi5Es,303
-gradientai/types/chat/completion_create_params.py,sha256=F4Dcrt5aYC_GEWUSN2OA3Zm5ImevZ7tFuTxBH7RyooM,6635
+gradientai/types/chat/__init__.py,sha256=c-PmEwuvWZQ4CRBTs9gzbKAq2sxL7V7JlVxddeoaGl0,381
+gradientai/types/chat/chat_completion_chunk.py,sha256=o1gDgYtzM477RmKWg-q5CE0tP3p0J7YKlZWaoqjCJOU,3046
+gradientai/types/chat/completion_create_params.py,sha256=ADEJ0N3MMsouT9AqBLE1-rho4FVVhlp9U9E-buqIAYs,7165
 gradientai/types/chat/completion_create_response.py,sha256=nNPWSXZYbyYLjT_ikVvDcjRw3f9eRGHFsUrLKtQHYGI,2413
 gradientai/types/inference/__init__.py,sha256=4Dt7-03NeP9ehdHLkLsZMiL_YLQwZsl92D0mMoDQ5g0,857
 gradientai/types/inference/api_key_create_params.py,sha256=MOy5Bdr1wNBqCvqzyZ0FLfFY2a97q6eXCzgCR1wcLAE,263
@@ -232,7 +233,7 @@ gradientai/types/shared/__init__.py,sha256=YA2_qLkZLySOac1HrqOfCTEz6GeipnjIJh1mK
 gradientai/types/shared/api_links.py,sha256=Iq5iQwOkRYuwLcuDLk54dUfrq0f2ZVEOXSpF744gYgA,403
 gradientai/types/shared/api_meta.py,sha256=-KyinzQqM5GSjD7E5xm7A4UALXAvLOyVNR1SYVOUFJM,297
 gradientai/types/shared/chat_completion_token_logprob.py,sha256=6-ipUFfsXMf5L7FDFi127NaVkDtmEooVgGBF6Ts965A,1769
-c63a5cfe_b235_4fbe_8bbb_82a9e02a482a_python-0.1.0a8.dist-info/METADATA,sha256=xwI-Z-rvG_Z_-L5Xu3x_NW2F7CauiUQmYG8wtunHk-Q,15049
-c63a5cfe_b235_4fbe_8bbb_82a9e02a482a_python-0.1.0a8.dist-info/WHEEL,sha256=C2FUgwZgiLbznR-k0b_5k3Ai_1aASOXDss3lzCUsUug,87
-c63a5cfe_b235_4fbe_8bbb_82a9e02a482a_python-0.1.0a8.dist-info/licenses/LICENSE,sha256=AzxEF8mEks6hu5V_87CXF8gLdL875WeO8FmQtEZTFok,11341
-c63a5cfe_b235_4fbe_8bbb_82a9e02a482a_python-0.1.0a8.dist-info/RECORD,,
+c63a5cfe_b235_4fbe_8bbb_82a9e02a482a_python-0.1.0a9.dist-info/METADATA,sha256=knvT5XUviLOr5lmtTU4M4N1a1_opUZDVrfNKXZM1nqM,15049
+c63a5cfe_b235_4fbe_8bbb_82a9e02a482a_python-0.1.0a9.dist-info/WHEEL,sha256=C2FUgwZgiLbznR-k0b_5k3Ai_1aASOXDss3lzCUsUug,87
+c63a5cfe_b235_4fbe_8bbb_82a9e02a482a_python-0.1.0a9.dist-info/licenses/LICENSE,sha256=AzxEF8mEks6hu5V_87CXF8gLdL875WeO8FmQtEZTFok,11341
+c63a5cfe_b235_4fbe_8bbb_82a9e02a482a_python-0.1.0a9.dist-info/RECORD,,

gradientai/_version.py

@@ -1,4 +1,4 @@
 # File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details.
 
 __title__ = "gradientai"
-__version__ = "0.1.0-alpha.8"  # x-release-please-version
+__version__ = "0.1.0-alpha.9"  # x-release-please-version

gradientai/resources/agents/chat/completions.py

@@ -20,8 +20,8 @@ from ...._response import (
 from ...._streaming import Stream, AsyncStream
 from ...._base_client import make_request_options
 from ....types.agents.chat import completion_create_params
-from ....types.agents.chat.chat_completion_chunk import ChatCompletionChunk
 from ....types.agents.chat.completion_create_response import CompletionCreateResponse
+from ....types.agents.chat.agent_chat_completion_chunk import AgentChatCompletionChunk
 
 __all__ = ["CompletionsResource", "AsyncCompletionsResource"]
 
@@ -186,7 +186,7 @@ class CompletionsResource(SyncAPIResource):
         extra_query: Query | None = None,
         extra_body: Body | None = None,
         timeout: float | httpx.Timeout | None | NotGiven = NOT_GIVEN,
-    ) -> Stream[ChatCompletionChunk]:
+    ) -> Stream[AgentChatCompletionChunk]:
         """
         Creates a model response for the given chat conversation.
 
@@ -299,7 +299,7 @@ class CompletionsResource(SyncAPIResource):
         extra_query: Query | None = None,
         extra_body: Body | None = None,
         timeout: float | httpx.Timeout | None | NotGiven = NOT_GIVEN,
-    ) -> CompletionCreateResponse | Stream[ChatCompletionChunk]:
+    ) -> CompletionCreateResponse | Stream[AgentChatCompletionChunk]:
         """
         Creates a model response for the given chat conversation.
 
@@ -412,7 +412,7 @@ class CompletionsResource(SyncAPIResource):
         extra_query: Query | None = None,
         extra_body: Body | None = None,
         timeout: float | httpx.Timeout | None | NotGiven = NOT_GIVEN,
-    ) -> CompletionCreateResponse | Stream[ChatCompletionChunk]:
+    ) -> CompletionCreateResponse | Stream[AgentChatCompletionChunk]:
         return self._post(
             "/chat/completions"
             if self._client._base_url_overridden
@@ -446,7 +446,7 @@ class CompletionsResource(SyncAPIResource):
             ),
             cast_to=CompletionCreateResponse,
             stream=stream or False,
-            stream_cls=Stream[ChatCompletionChunk],
+            stream_cls=Stream[AgentChatCompletionChunk],
         )
 
 
@@ -610,7 +610,7 @@ class AsyncCompletionsResource(AsyncAPIResource):
         extra_query: Query | None = None,
         extra_body: Body | None = None,
         timeout: float | httpx.Timeout | None | NotGiven = NOT_GIVEN,
-    ) -> AsyncStream[ChatCompletionChunk]:
+    ) -> AsyncStream[AgentChatCompletionChunk]:
         """
         Creates a model response for the given chat conversation.
 
@@ -723,7 +723,7 @@ class AsyncCompletionsResource(AsyncAPIResource):
         extra_query: Query | None = None,
         extra_body: Body | None = None,
         timeout: float | httpx.Timeout | None | NotGiven = NOT_GIVEN,
-    ) -> CompletionCreateResponse | AsyncStream[ChatCompletionChunk]:
+    ) -> CompletionCreateResponse | AsyncStream[AgentChatCompletionChunk]:
         """
         Creates a model response for the given chat conversation.
 
@@ -836,7 +836,7 @@ class AsyncCompletionsResource(AsyncAPIResource):
         extra_query: Query | None = None,
         extra_body: Body | None = None,
         timeout: float | httpx.Timeout | None | NotGiven = NOT_GIVEN,
-    ) -> CompletionCreateResponse | AsyncStream[ChatCompletionChunk]:
+    ) -> CompletionCreateResponse | AsyncStream[AgentChatCompletionChunk]:
         return await self._post(
             "/chat/completions"
             if self._client._base_url_overridden
@@ -870,7 +870,7 @@ class AsyncCompletionsResource(AsyncAPIResource):
             ),
             cast_to=CompletionCreateResponse,
             stream=stream or False,
-            stream_cls=AsyncStream[ChatCompletionChunk],
+            stream_cls=AsyncStream[AgentChatCompletionChunk],
         )
 
 

gradientai/resources/chat/completions.py

@@ -3,11 +3,12 @@
 from __future__ import annotations
 
 from typing import Dict, List, Union, Iterable, Optional
+from typing_extensions import Literal, overload
 
 import httpx
 
 from ..._types import NOT_GIVEN, Body, Query, Headers, NotGiven
-from ..._utils import maybe_transform, async_maybe_transform
+from ..._utils import required_args, maybe_transform, async_maybe_transform
 from ..._compat import cached_property
 from ..._resource import SyncAPIResource, AsyncAPIResource
 from ..._response import (
@@ -16,8 +17,10 @@ from ..._response import (
     async_to_raw_response_wrapper,
     async_to_streamed_response_wrapper,
 )
+from ..._streaming import Stream, AsyncStream
 from ...types.chat import completion_create_params
 from ..._base_client import make_request_options
+from ...types.chat.chat_completion_chunk import ChatCompletionChunk
 from ...types.chat.completion_create_response import CompletionCreateResponse
 
 __all__ = ["CompletionsResource", "AsyncCompletionsResource"]
@@ -43,6 +46,7 @@ class CompletionsResource(SyncAPIResource):
         """
         return CompletionsResourceWithStreamingResponse(self)
 
+    @overload
     def create(
         self,
         *,
@@ -57,7 +61,7 @@ class CompletionsResource(SyncAPIResource):
         n: Optional[int] | NotGiven = NOT_GIVEN,
         presence_penalty: Optional[float] | NotGiven = NOT_GIVEN,
         stop: Union[Optional[str], List[str], None] | NotGiven = NOT_GIVEN,
-        stream: Optional[bool] | NotGiven = NOT_GIVEN,
+        stream: Optional[Literal[False]] | NotGiven = NOT_GIVEN,
         stream_options: Optional[completion_create_params.StreamOptions] | NotGiven = NOT_GIVEN,
         temperature: Optional[float] | NotGiven = NOT_GIVEN,
         top_logprobs: Optional[int] | NotGiven = NOT_GIVEN,
@@ -154,6 +158,263 @@ class CompletionsResource(SyncAPIResource):
           timeout: Override the client-level default timeout for this request, in seconds
         """
 
+        ...
+
+    @overload
+    def create(
+        self,
+        *,
+        messages: Iterable[completion_create_params.Message],
+        model: str,
+        stream: Literal[True],
+        frequency_penalty: Optional[float] | NotGiven = NOT_GIVEN,
+        logit_bias: Optional[Dict[str, int]] | NotGiven = NOT_GIVEN,
+        logprobs: Optional[bool] | NotGiven = NOT_GIVEN,
+        max_completion_tokens: Optional[int] | NotGiven = NOT_GIVEN,
+        max_tokens: Optional[int] | NotGiven = NOT_GIVEN,
+        metadata: Optional[Dict[str, str]] | NotGiven = NOT_GIVEN,
+        n: Optional[int] | NotGiven = NOT_GIVEN,
+        presence_penalty: Optional[float] | NotGiven = NOT_GIVEN,
+        stop: Union[Optional[str], List[str], None] | NotGiven = NOT_GIVEN,
+        stream_options: Optional[completion_create_params.StreamOptions] | NotGiven = NOT_GIVEN,
+        temperature: Optional[float] | NotGiven = NOT_GIVEN,
+        top_logprobs: Optional[int] | NotGiven = NOT_GIVEN,
+        top_p: Optional[float] | NotGiven = NOT_GIVEN,
+        user: str | 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,
+        extra_query: Query | None = None,
+        extra_body: Body | None = None,
+        timeout: float | httpx.Timeout | None | NotGiven = NOT_GIVEN,
+    ) -> Stream[ChatCompletionChunk]:
+        """
+        Creates a model response for the given chat conversation.
+
+        Args:
+          messages: A list of messages comprising the conversation so far.
+
+          model: Model ID used to generate the response.
+
+          stream: If set to true, the model response data will be streamed to the client as it is
+              generated using server-sent events.
+
+          frequency_penalty: Number between -2.0 and 2.0. Positive values penalize new tokens based on their
+              existing frequency in the text so far, decreasing the model's likelihood to
+              repeat the same line verbatim.
+
+          logit_bias: Modify the likelihood of specified tokens appearing in the completion.
+
+              Accepts a JSON object that maps tokens (specified by their token ID in the
+              tokenizer) to an associated bias value from -100 to 100. Mathematically, the
+              bias is added to the logits generated by the model prior to sampling. The exact
+              effect will vary per model, but values between -1 and 1 should decrease or
+              increase likelihood of selection; values like -100 or 100 should result in a ban
+              or exclusive selection of the relevant token.
+
+          logprobs: Whether to return log probabilities of the output tokens or not. If true,
+              returns the log probabilities of each output token returned in the `content` of
+              `message`.
+
+          max_completion_tokens: The maximum number of completion tokens that may be used over the course of the
+              run. The run will make a best effort to use only the number of completion tokens
+              specified, across multiple turns of the run.
+
+          max_tokens: The maximum number of tokens that can be generated in the completion.
+
+              The token count of your prompt plus `max_tokens` cannot exceed the model's
+              context length.
+
+          metadata: Set of 16 key-value pairs that can be attached to an object. This can be useful
+              for storing additional information about the object in a structured format, and
+              querying for objects via API or the dashboard.
+
+              Keys are strings with a maximum length of 64 characters. Values are strings with
+              a maximum length of 512 characters.
+
+          n: How many chat completion choices to generate for each input message. Note that
+              you will be charged based on the number of generated tokens across all of the
+              choices. Keep `n` as `1` to minimize costs.
+
+          presence_penalty: Number between -2.0 and 2.0. Positive values penalize new tokens based on
+              whether they appear in the text so far, increasing the model's likelihood to
+              talk about new topics.
+
+          stop: Up to 4 sequences where the API will stop generating further tokens. The
+              returned text will not contain the stop sequence.
+
+          stream_options: Options for streaming response. Only set this when you set `stream: true`.
+
+          temperature: What sampling temperature to use, between 0 and 2. Higher values like 0.8 will
+              make the output more random, while lower values like 0.2 will make it more
+              focused and deterministic. We generally recommend altering this or `top_p` but
+              not both.
+
+          top_logprobs: An integer between 0 and 20 specifying the number of most likely tokens to
+              return at each token position, each with an associated log probability.
+              `logprobs` must be set to `true` if this parameter is used.
+
+          top_p: An alternative to sampling with temperature, called nucleus sampling, where the
+              model considers the results of the tokens with top_p probability mass. So 0.1
+              means only the tokens comprising the top 10% probability mass are considered.
+
+              We generally recommend altering this or `temperature` but not both.
+
+          user: A unique identifier representing your end-user, which can help DigitalOcean to
+              monitor and detect abuse.
+
+          extra_headers: Send extra headers
+
+          extra_query: Add additional query parameters to the request
+
+          extra_body: Add additional JSON properties to the request
+
+          timeout: Override the client-level default timeout for this request, in seconds
+        """
+        ...
+
+    @overload
+    def create(
+        self,
+        *,
+        messages: Iterable[completion_create_params.Message],
+        model: str,
+        stream: bool,
+        frequency_penalty: Optional[float] | NotGiven = NOT_GIVEN,
+        logit_bias: Optional[Dict[str, int]] | NotGiven = NOT_GIVEN,
+        logprobs: Optional[bool] | NotGiven = NOT_GIVEN,
+        max_completion_tokens: Optional[int] | NotGiven = NOT_GIVEN,
+        max_tokens: Optional[int] | NotGiven = NOT_GIVEN,
+        metadata: Optional[Dict[str, str]] | NotGiven = NOT_GIVEN,
+        n: Optional[int] | NotGiven = NOT_GIVEN,
+        presence_penalty: Optional[float] | NotGiven = NOT_GIVEN,
+        stop: Union[Optional[str], List[str], None] | NotGiven = NOT_GIVEN,
+        stream_options: Optional[completion_create_params.StreamOptions] | NotGiven = NOT_GIVEN,
+        temperature: Optional[float] | NotGiven = NOT_GIVEN,
+        top_logprobs: Optional[int] | NotGiven = NOT_GIVEN,
+        top_p: Optional[float] | NotGiven = NOT_GIVEN,
+        user: str | 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,
+        extra_query: Query | None = None,
+        extra_body: Body | None = None,
+        timeout: float | httpx.Timeout | None | NotGiven = NOT_GIVEN,
+    ) -> CompletionCreateResponse | Stream[ChatCompletionChunk]:
+        """
+        Creates a model response for the given chat conversation.
+
+        Args:
+          messages: A list of messages comprising the conversation so far.
+
+          model: Model ID used to generate the response.
+
+          stream: If set to true, the model response data will be streamed to the client as it is
+              generated using server-sent events.
+
+          frequency_penalty: Number between -2.0 and 2.0. Positive values penalize new tokens based on their
+              existing frequency in the text so far, decreasing the model's likelihood to
+              repeat the same line verbatim.
+
+          logit_bias: Modify the likelihood of specified tokens appearing in the completion.
+
+              Accepts a JSON object that maps tokens (specified by their token ID in the
+              tokenizer) to an associated bias value from -100 to 100. Mathematically, the
+              bias is added to the logits generated by the model prior to sampling. The exact
+              effect will vary per model, but values between -1 and 1 should decrease or
+              increase likelihood of selection; values like -100 or 100 should result in a ban
+              or exclusive selection of the relevant token.
+
+          logprobs: Whether to return log probabilities of the output tokens or not. If true,
+              returns the log probabilities of each output token returned in the `content` of
+              `message`.
+
+          max_completion_tokens: The maximum number of completion tokens that may be used over the course of the
+              run. The run will make a best effort to use only the number of completion tokens
+              specified, across multiple turns of the run.
+
+          max_tokens: The maximum number of tokens that can be generated in the completion.
+
+              The token count of your prompt plus `max_tokens` cannot exceed the model's
+              context length.
+
+          metadata: Set of 16 key-value pairs that can be attached to an object. This can be useful
+              for storing additional information about the object in a structured format, and
+              querying for objects via API or the dashboard.
+
+              Keys are strings with a maximum length of 64 characters. Values are strings with
+              a maximum length of 512 characters.
+
+          n: How many chat completion choices to generate for each input message. Note that
+              you will be charged based on the number of generated tokens across all of the
+              choices. Keep `n` as `1` to minimize costs.
+
+          presence_penalty: Number between -2.0 and 2.0. Positive values penalize new tokens based on
+              whether they appear in the text so far, increasing the model's likelihood to
+              talk about new topics.
+
+          stop: Up to 4 sequences where the API will stop generating further tokens. The
+              returned text will not contain the stop sequence.
+
+          stream_options: Options for streaming response. Only set this when you set `stream: true`.
+
+          temperature: What sampling temperature to use, between 0 and 2. Higher values like 0.8 will
+              make the output more random, while lower values like 0.2 will make it more
+              focused and deterministic. We generally recommend altering this or `top_p` but
+              not both.
+
+          top_logprobs: An integer between 0 and 20 specifying the number of most likely tokens to
+              return at each token position, each with an associated log probability.
+              `logprobs` must be set to `true` if this parameter is used.
+
+          top_p: An alternative to sampling with temperature, called nucleus sampling, where the
+              model considers the results of the tokens with top_p probability mass. So 0.1
+              means only the tokens comprising the top 10% probability mass are considered.
+
+              We generally recommend altering this or `temperature` but not both.
+
+          user: A unique identifier representing your end-user, which can help DigitalOcean to
+              monitor and detect abuse.
+
+          extra_headers: Send extra headers
+
+          extra_query: Add additional query parameters to the request
+
+          extra_body: Add additional JSON properties to the request
+
+          timeout: Override the client-level default timeout for this request, in seconds
+        """
+        ...
+
+    @required_args(["messages", "model"], ["messages", "model", "stream"])
+    def create(
+        self,
+        *,
+        messages: Iterable[completion_create_params.Message],
+        model: str,
+        frequency_penalty: Optional[float] | NotGiven = NOT_GIVEN,
+        logit_bias: Optional[Dict[str, int]] | NotGiven = NOT_GIVEN,
+        logprobs: Optional[bool] | NotGiven = NOT_GIVEN,
+        max_completion_tokens: Optional[int] | NotGiven = NOT_GIVEN,
+        max_tokens: Optional[int] | NotGiven = NOT_GIVEN,
+        metadata: Optional[Dict[str, str]] | NotGiven = NOT_GIVEN,
+        n: Optional[int] | NotGiven = NOT_GIVEN,
+        presence_penalty: Optional[float] | NotGiven = NOT_GIVEN,
+        stop: Union[Optional[str], List[str], None] | NotGiven = NOT_GIVEN,
+        stream: Optional[Literal[False]] | Literal[True] | NotGiven = NOT_GIVEN,
+        stream_options: Optional[completion_create_params.StreamOptions] | NotGiven = NOT_GIVEN,
+        temperature: Optional[float] | NotGiven = NOT_GIVEN,
+        top_logprobs: Optional[int] | NotGiven = NOT_GIVEN,
+        top_p: Optional[float] | NotGiven = NOT_GIVEN,
+        user: str | 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,
+        extra_query: Query | None = None,
+        extra_body: Body | None = None,
+        timeout: float | httpx.Timeout | None | NotGiven = NOT_GIVEN,
+    ) -> CompletionCreateResponse | Stream[ChatCompletionChunk]:
+
         # This method requires an inference_key to be set via client argument or environment variable
         if not self._client.inference_key:
             raise TypeError(
@@ -186,12 +447,16 @@ class CompletionsResource(SyncAPIResource):
                     "top_p": top_p,
                     "user": user,
                 },
-                completion_create_params.CompletionCreateParams,
+                completion_create_params.CompletionCreateParamsStreaming
+                if stream
+                else completion_create_params.CompletionCreateParamsNonStreaming,
             ),
             options=make_request_options(
                 extra_headers=headers, extra_query=extra_query, extra_body=extra_body, timeout=timeout
             ),
             cast_to=CompletionCreateResponse,
+            stream=stream or False,
+            stream_cls=Stream[ChatCompletionChunk],
         )
 
 
@@ -215,6 +480,7 @@ class AsyncCompletionsResource(AsyncAPIResource):
         """
         return AsyncCompletionsResourceWithStreamingResponse(self)
 
+    @overload
     async def create(
         self,
         *,
@@ -229,7 +495,7 @@ class AsyncCompletionsResource(AsyncAPIResource):
         n: Optional[int] | NotGiven = NOT_GIVEN,
         presence_penalty: Optional[float] | NotGiven = NOT_GIVEN,
         stop: Union[Optional[str], List[str], None] | NotGiven = NOT_GIVEN,
-        stream: Optional[bool] | NotGiven = NOT_GIVEN,
+        stream: Optional[Literal[False]] | NotGiven = NOT_GIVEN,
         stream_options: Optional[completion_create_params.StreamOptions] | NotGiven = NOT_GIVEN,
         temperature: Optional[float] | NotGiven = NOT_GIVEN,
         top_logprobs: Optional[int] | NotGiven = NOT_GIVEN,
@@ -326,6 +592,263 @@ class AsyncCompletionsResource(AsyncAPIResource):
           timeout: Override the client-level default timeout for this request, in seconds
         """
 
+        ...
+
+    @overload
+    async def create(
+        self,
+        *,
+        messages: Iterable[completion_create_params.Message],
+        model: str,
+        stream: Literal[True],
+        frequency_penalty: Optional[float] | NotGiven = NOT_GIVEN,
+        logit_bias: Optional[Dict[str, int]] | NotGiven = NOT_GIVEN,
+        logprobs: Optional[bool] | NotGiven = NOT_GIVEN,
+        max_completion_tokens: Optional[int] | NotGiven = NOT_GIVEN,
+        max_tokens: Optional[int] | NotGiven = NOT_GIVEN,
+        metadata: Optional[Dict[str, str]] | NotGiven = NOT_GIVEN,
+        n: Optional[int] | NotGiven = NOT_GIVEN,
+        presence_penalty: Optional[float] | NotGiven = NOT_GIVEN,
+        stop: Union[Optional[str], List[str], None] | NotGiven = NOT_GIVEN,
+        stream_options: Optional[completion_create_params.StreamOptions] | NotGiven = NOT_GIVEN,
+        temperature: Optional[float] | NotGiven = NOT_GIVEN,
+        top_logprobs: Optional[int] | NotGiven = NOT_GIVEN,
+        top_p: Optional[float] | NotGiven = NOT_GIVEN,
+        user: str | 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,
+        extra_query: Query | None = None,
+        extra_body: Body | None = None,
+        timeout: float | httpx.Timeout | None | NotGiven = NOT_GIVEN,
+    ) -> AsyncStream[ChatCompletionChunk]:
+        """
+        Creates a model response for the given chat conversation.
+
+        Args:
+          messages: A list of messages comprising the conversation so far.
+
+          model: Model ID used to generate the response.
+
+          stream: If set to true, the model response data will be streamed to the client as it is
+              generated using server-sent events.
+
+          frequency_penalty: Number between -2.0 and 2.0. Positive values penalize new tokens based on their
+              existing frequency in the text so far, decreasing the model's likelihood to
+              repeat the same line verbatim.
+
+          logit_bias: Modify the likelihood of specified tokens appearing in the completion.
+
+              Accepts a JSON object that maps tokens (specified by their token ID in the
+              tokenizer) to an associated bias value from -100 to 100. Mathematically, the
+              bias is added to the logits generated by the model prior to sampling. The exact
+              effect will vary per model, but values between -1 and 1 should decrease or
+              increase likelihood of selection; values like -100 or 100 should result in a ban
+              or exclusive selection of the relevant token.
+
+          logprobs: Whether to return log probabilities of the output tokens or not. If true,
+              returns the log probabilities of each output token returned in the `content` of
+              `message`.
+
+          max_completion_tokens: The maximum number of completion tokens that may be used over the course of the
+              run. The run will make a best effort to use only the number of completion tokens
+              specified, across multiple turns of the run.
+
+          max_tokens: The maximum number of tokens that can be generated in the completion.
+
+              The token count of your prompt plus `max_tokens` cannot exceed the model's
+              context length.
+
+          metadata: Set of 16 key-value pairs that can be attached to an object. This can be useful
+              for storing additional information about the object in a structured format, and
+              querying for objects via API or the dashboard.
+
+              Keys are strings with a maximum length of 64 characters. Values are strings with
+              a maximum length of 512 characters.
+
+          n: How many chat completion choices to generate for each input message. Note that
+              you will be charged based on the number of generated tokens across all of the
+              choices. Keep `n` as `1` to minimize costs.
+
+          presence_penalty: Number between -2.0 and 2.0. Positive values penalize new tokens based on
+              whether they appear in the text so far, increasing the model's likelihood to
+              talk about new topics.
+
+          stop: Up to 4 sequences where the API will stop generating further tokens. The
+              returned text will not contain the stop sequence.
+
+          stream_options: Options for streaming response. Only set this when you set `stream: true`.
+
+          temperature: What sampling temperature to use, between 0 and 2. Higher values like 0.8 will
+              make the output more random, while lower values like 0.2 will make it more
+              focused and deterministic. We generally recommend altering this or `top_p` but
+              not both.
+
+          top_logprobs: An integer between 0 and 20 specifying the number of most likely tokens to
+              return at each token position, each with an associated log probability.
+              `logprobs` must be set to `true` if this parameter is used.
+
+          top_p: An alternative to sampling with temperature, called nucleus sampling, where the
+              model considers the results of the tokens with top_p probability mass. So 0.1
+              means only the tokens comprising the top 10% probability mass are considered.
+
+              We generally recommend altering this or `temperature` but not both.
+
+          user: A unique identifier representing your end-user, which can help DigitalOcean to
+              monitor and detect abuse.
+
+          extra_headers: Send extra headers
+
+          extra_query: Add additional query parameters to the request
+
+          extra_body: Add additional JSON properties to the request
+
+          timeout: Override the client-level default timeout for this request, in seconds
+        """
+        ...
+
+    @overload
+    async def create(
+        self,
+        *,
+        messages: Iterable[completion_create_params.Message],
+        model: str,
+        stream: bool,
+        frequency_penalty: Optional[float] | NotGiven = NOT_GIVEN,
+        logit_bias: Optional[Dict[str, int]] | NotGiven = NOT_GIVEN,
+        logprobs: Optional[bool] | NotGiven = NOT_GIVEN,
+        max_completion_tokens: Optional[int] | NotGiven = NOT_GIVEN,
+        max_tokens: Optional[int] | NotGiven = NOT_GIVEN,
+        metadata: Optional[Dict[str, str]] | NotGiven = NOT_GIVEN,
+        n: Optional[int] | NotGiven = NOT_GIVEN,
+        presence_penalty: Optional[float] | NotGiven = NOT_GIVEN,
+        stop: Union[Optional[str], List[str], None] | NotGiven = NOT_GIVEN,
+        stream_options: Optional[completion_create_params.StreamOptions] | NotGiven = NOT_GIVEN,
+        temperature: Optional[float] | NotGiven = NOT_GIVEN,
+        top_logprobs: Optional[int] | NotGiven = NOT_GIVEN,
+        top_p: Optional[float] | NotGiven = NOT_GIVEN,
+        user: str | 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,
+        extra_query: Query | None = None,
+        extra_body: Body | None = None,
+        timeout: float | httpx.Timeout | None | NotGiven = NOT_GIVEN,
+    ) -> CompletionCreateResponse | AsyncStream[ChatCompletionChunk]:
+        """
+        Creates a model response for the given chat conversation.
+
+        Args:
+          messages: A list of messages comprising the conversation so far.
+
+          model: Model ID used to generate the response.
+
+          stream: If set to true, the model response data will be streamed to the client as it is
+              generated using server-sent events.
+
+          frequency_penalty: Number between -2.0 and 2.0. Positive values penalize new tokens based on their
+              existing frequency in the text so far, decreasing the model's likelihood to
+              repeat the same line verbatim.
+
+          logit_bias: Modify the likelihood of specified tokens appearing in the completion.
+
+              Accepts a JSON object that maps tokens (specified by their token ID in the
+              tokenizer) to an associated bias value from -100 to 100. Mathematically, the
+              bias is added to the logits generated by the model prior to sampling. The exact
+              effect will vary per model, but values between -1 and 1 should decrease or
+              increase likelihood of selection; values like -100 or 100 should result in a ban
+              or exclusive selection of the relevant token.
+
+          logprobs: Whether to return log probabilities of the output tokens or not. If true,
+              returns the log probabilities of each output token returned in the `content` of
+              `message`.
+
+          max_completion_tokens: The maximum number of completion tokens that may be used over the course of the
+              run. The run will make a best effort to use only the number of completion tokens
+              specified, across multiple turns of the run.
+
+          max_tokens: The maximum number of tokens that can be generated in the completion.
+
+              The token count of your prompt plus `max_tokens` cannot exceed the model's
+              context length.
+
+          metadata: Set of 16 key-value pairs that can be attached to an object. This can be useful
+              for storing additional information about the object in a structured format, and
+              querying for objects via API or the dashboard.
+
+              Keys are strings with a maximum length of 64 characters. Values are strings with
+              a maximum length of 512 characters.
+
+          n: How many chat completion choices to generate for each input message. Note that
+              you will be charged based on the number of generated tokens across all of the
+              choices. Keep `n` as `1` to minimize costs.
+
+          presence_penalty: Number between -2.0 and 2.0. Positive values penalize new tokens based on
+              whether they appear in the text so far, increasing the model's likelihood to
+              talk about new topics.
+
+          stop: Up to 4 sequences where the API will stop generating further tokens. The
+              returned text will not contain the stop sequence.
+
+          stream_options: Options for streaming response. Only set this when you set `stream: true`.
+
+          temperature: What sampling temperature to use, between 0 and 2. Higher values like 0.8 will
+              make the output more random, while lower values like 0.2 will make it more
+              focused and deterministic. We generally recommend altering this or `top_p` but
+              not both.
+
+          top_logprobs: An integer between 0 and 20 specifying the number of most likely tokens to
+              return at each token position, each with an associated log probability.
+              `logprobs` must be set to `true` if this parameter is used.
+
+          top_p: An alternative to sampling with temperature, called nucleus sampling, where the
+              model considers the results of the tokens with top_p probability mass. So 0.1
+              means only the tokens comprising the top 10% probability mass are considered.
+
+              We generally recommend altering this or `temperature` but not both.
+
+          user: A unique identifier representing your end-user, which can help DigitalOcean to
+              monitor and detect abuse.
+
+          extra_headers: Send extra headers
+
+          extra_query: Add additional query parameters to the request
+
+          extra_body: Add additional JSON properties to the request
+
+          timeout: Override the client-level default timeout for this request, in seconds
+        """
+        ...
+
+    @required_args(["messages", "model"], ["messages", "model", "stream"])
+    async def create(
+        self,
+        *,
+        messages: Iterable[completion_create_params.Message],
+        model: str,
+        frequency_penalty: Optional[float] | NotGiven = NOT_GIVEN,
+        logit_bias: Optional[Dict[str, int]] | NotGiven = NOT_GIVEN,
+        logprobs: Optional[bool] | NotGiven = NOT_GIVEN,
+        max_completion_tokens: Optional[int] | NotGiven = NOT_GIVEN,
+        max_tokens: Optional[int] | NotGiven = NOT_GIVEN,
+        metadata: Optional[Dict[str, str]] | NotGiven = NOT_GIVEN,
+        n: Optional[int] | NotGiven = NOT_GIVEN,
+        presence_penalty: Optional[float] | NotGiven = NOT_GIVEN,
+        stop: Union[Optional[str], List[str], None] | NotGiven = NOT_GIVEN,
+        stream: Optional[Literal[False]] | Literal[True] | NotGiven = NOT_GIVEN,
+        stream_options: Optional[completion_create_params.StreamOptions] | NotGiven = NOT_GIVEN,
+        temperature: Optional[float] | NotGiven = NOT_GIVEN,
+        top_logprobs: Optional[int] | NotGiven = NOT_GIVEN,
+        top_p: Optional[float] | NotGiven = NOT_GIVEN,
+        user: str | 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,
+        extra_query: Query | None = None,
+        extra_body: Body | None = None,
+        timeout: float | httpx.Timeout | None | NotGiven = NOT_GIVEN,
+    ) -> CompletionCreateResponse | AsyncStream[ChatCompletionChunk]:
+
         # This method requires an inference_key to be set via client argument or environment variable
         if not hasattr(self._client, "inference_key") or not self._client.inference_key:
             raise TypeError(
@@ -358,12 +881,16 @@ class AsyncCompletionsResource(AsyncAPIResource):
                     "top_p": top_p,
                     "user": user,
                 },
-                completion_create_params.CompletionCreateParams,
+                completion_create_params.CompletionCreateParamsStreaming
+                if stream
+                else completion_create_params.CompletionCreateParamsNonStreaming,
             ),
             options=make_request_options(
                 extra_headers=headers, extra_query=extra_query, extra_body=extra_body, timeout=timeout
             ),
             cast_to=CompletionCreateResponse,
+            stream=stream or False,
+            stream_cls=AsyncStream[ChatCompletionChunk],
         )
 
 

gradientai/types/agents/chat/__init__.py

@@ -2,6 +2,6 @@
 
 from __future__ import annotations
 
-from .chat_completion_chunk import ChatCompletionChunk as ChatCompletionChunk
 from .completion_create_params import CompletionCreateParams as CompletionCreateParams
 from .completion_create_response import CompletionCreateResponse as CompletionCreateResponse
+from .agent_chat_completion_chunk import AgentChatCompletionChunk as AgentChatCompletionChunk

gradientai/types/agents/chat/agent_chat_completion_chunk.py

@@ -0,0 +1,93 @@
+# 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 ...shared.chat_completion_token_logprob import ChatCompletionTokenLogprob
+
+__all__ = ["AgentChatCompletionChunk", "Choice", "ChoiceDelta", "ChoiceLogprobs", "Usage"]
+
+
+class ChoiceDelta(BaseModel):
+    content: Optional[str] = None
+    """The contents of the chunk message."""
+
+    refusal: Optional[str] = None
+    """The refusal message generated by the model."""
+
+    role: Optional[Literal["developer", "user", "assistant"]] = None
+    """The role of the author of this message."""
+
+
+class ChoiceLogprobs(BaseModel):
+    content: Optional[List[ChatCompletionTokenLogprob]] = None
+    """A list of message content tokens with log probability information."""
+
+    refusal: Optional[List[ChatCompletionTokenLogprob]] = None
+    """A list of message refusal tokens with log probability information."""
+
+
+class Choice(BaseModel):
+    delta: ChoiceDelta
+    """A chat completion delta generated by streamed model responses."""
+
+    finish_reason: Optional[Literal["stop", "length"]] = None
+    """The reason the model stopped generating tokens.
+
+    This will be `stop` if the model hit a natural stop point or a provided stop
+    sequence, or `length` if the maximum number of tokens specified in the request
+    was reached
+    """
+
+    index: int
+    """The index of the choice in the list of choices."""
+
+    logprobs: Optional[ChoiceLogprobs] = None
+    """Log probability information for the choice."""
+
+
+class Usage(BaseModel):
+    completion_tokens: int
+    """Number of tokens in the generated completion."""
+
+    prompt_tokens: int
+    """Number of tokens in the prompt."""
+
+    total_tokens: int
+    """Total number of tokens used in the request (prompt + completion)."""
+
+
+class AgentChatCompletionChunk(BaseModel):
+    id: str
+    """A unique identifier for the chat completion. Each chunk has the same ID."""
+
+    choices: List[Choice]
+    """A list of chat completion choices.
+
+    Can contain more than one elements if `n` is greater than 1. Can also be empty
+    for the last chunk if you set `stream_options: {"include_usage": true}`.
+    """
+
+    created: int
+    """The Unix timestamp (in seconds) of when the chat completion was created.
+
+    Each chunk has the same timestamp.
+    """
+
+    model: str
+    """The model to generate the completion."""
+
+    object: Literal["chat.completion.chunk"]
+    """The object type, which is always `chat.completion.chunk`."""
+
+    usage: Optional[Usage] = None
+    """
+    An optional field that will only be present when you set
+    `stream_options: {"include_usage": true}` in your request. When present, it
+    contains a null value **except for the last chunk** which contains the token
+    usage statistics for the entire request.
+
+    **NOTE:** If the stream is interrupted or cancelled, you may not receive the
+    final usage chunk which contains the total token usage for the request.
+    """

gradientai/types/chat/__init__.py

@@ -2,5 +2,6 @@
 
 from __future__ import annotations
 
+from .chat_completion_chunk import ChatCompletionChunk as ChatCompletionChunk
 from .completion_create_params import CompletionCreateParams as CompletionCreateParams
 from .completion_create_response import CompletionCreateResponse as CompletionCreateResponse

gradientai/types/{agents/chat → chat}/chat_completion_chunk.py

@@ -3,8 +3,8 @@
 from typing import List, Optional
 from typing_extensions import Literal
 
-from ...._models import BaseModel
-from ...shared.chat_completion_token_logprob import ChatCompletionTokenLogprob
+from ..._models import BaseModel
+from ..shared.chat_completion_token_logprob import ChatCompletionTokenLogprob
 
 __all__ = ["ChatCompletionChunk", "Choice", "ChoiceDelta", "ChoiceLogprobs", "Usage"]
 

gradientai/types/chat/completion_create_params.py

@@ -6,17 +6,19 @@ from typing import Dict, List, Union, Iterable, Optional
 from typing_extensions import Literal, Required, TypeAlias, TypedDict
 
 __all__ = [
-    "CompletionCreateParams",
+    "CompletionCreateParamsBase",
     "Message",
     "MessageChatCompletionRequestSystemMessage",
     "MessageChatCompletionRequestDeveloperMessage",
     "MessageChatCompletionRequestUserMessage",
     "MessageChatCompletionRequestAssistantMessage",
     "StreamOptions",
+    "CompletionCreateParamsNonStreaming",
+    "CompletionCreateParamsStreaming",
 ]
 
 
-class CompletionCreateParams(TypedDict, total=False):
+class CompletionCreateParamsBase(TypedDict, total=False):
     messages: Required[Iterable[Message]]
     """A list of messages comprising the conversation so far."""
 
@@ -92,12 +94,6 @@ class CompletionCreateParams(TypedDict, total=False):
     The returned text will not contain the stop sequence.
     """
 
-    stream: Optional[bool]
-    """
-    If set to true, the model response data will be streamed to the client as it is
-    generated using server-sent events.
-    """
-
     stream_options: Optional[StreamOptions]
     """Options for streaming response. Only set this when you set `stream: true`."""
 
@@ -183,3 +179,22 @@ class StreamOptions(TypedDict, total=False):
     **NOTE:** If the stream is interrupted, you may not receive the final usage
     chunk which contains the total token usage for the request.
     """
+
+
+class CompletionCreateParamsNonStreaming(CompletionCreateParamsBase, total=False):
+    stream: Optional[Literal[False]]
+    """
+    If set to true, the model response data will be streamed to the client as it is
+    generated using server-sent events.
+    """
+
+
+class CompletionCreateParamsStreaming(CompletionCreateParamsBase):
+    stream: Required[Literal[True]]
+    """
+    If set to true, the model response data will be streamed to the client as it is
+    generated using server-sent events.
+    """
+
+
+CompletionCreateParams = Union[CompletionCreateParamsNonStreaming, CompletionCreateParamsStreaming]