# This file was auto-generated by Fern from our API Definition. import typing from json.decoder import JSONDecodeError from ..core.api_error import ApiError from ..core.client_wrapper import AsyncClientWrapper, SyncClientWrapper from ..core.http_response import AsyncHttpResponse, HttpResponse from ..core.pydantic_utilities import parse_obj_as from ..core.request_options import RequestOptions from ..core.serialization import convert_and_respect_annotation_metadata from ..types.chat_completion_response import ChatCompletionResponse from ..types.function import Function from ..types.function_call import FunctionCall from ..types.message import Message from ..types.response_format import ResponseFormat from ..types.stream_options import StreamOptions from ..types.tool import Tool # this is used as the default value for optional parameters OMIT = typing.cast(typing.Any, ...) class RawChatCompletionsClient: def __init__(self, *, client_wrapper: SyncClientWrapper): self._client_wrapper = client_wrapper def create_chat_completions( self, *, model: typing.Optional[str] = OMIT, messages: typing.Optional[typing.Sequence[Message]] = OMIT, temperature: typing.Optional[float] = OMIT, top_p: typing.Optional[float] = OMIT, n: typing.Optional[int] = OMIT, stream: typing.Optional[bool] = OMIT, stream_options: typing.Optional[StreamOptions] = OMIT, stop: typing.Optional[typing.Sequence[str]] = OMIT, max_tokens: typing.Optional[int] = OMIT, max_completion_tokens: typing.Optional[int] = OMIT, presence_penalty: typing.Optional[float] = OMIT, frequency_penalty: typing.Optional[float] = OMIT, logit_bias: typing.Optional[typing.Dict[str, int]] = OMIT, user: typing.Optional[str] = OMIT, response_format: typing.Optional[ResponseFormat] = OMIT, seed: typing.Optional[int] = OMIT, tools: typing.Optional[typing.Sequence[Tool]] = OMIT, tool_choice: typing.Optional[typing.Dict[str, typing.Optional[typing.Any]]] = OMIT, parallel_tool_calls: typing.Optional[bool] = OMIT, store: typing.Optional[bool] = OMIT, metadata: typing.Optional[typing.Dict[str, str]] = OMIT, reasoning_effort: typing.Optional[str] = OMIT, service_tier: typing.Optional[str] = OMIT, functions: typing.Optional[typing.Sequence[Function]] = OMIT, function_call: typing.Optional[FunctionCall] = OMIT, request_options: typing.Optional[RequestOptions] = None, ) -> HttpResponse[ChatCompletionResponse]: """ Create chat completions Parameters ---------- model : typing.Optional[str] messages : typing.Optional[typing.Sequence[Message]] temperature : typing.Optional[float] top_p : typing.Optional[float] n : typing.Optional[int] stream : typing.Optional[bool] stream_options : typing.Optional[StreamOptions] stop : typing.Optional[typing.Sequence[str]] max_tokens : typing.Optional[int] max_completion_tokens : typing.Optional[int] presence_penalty : typing.Optional[float] frequency_penalty : typing.Optional[float] logit_bias : typing.Optional[typing.Dict[str, int]] user : typing.Optional[str] response_format : typing.Optional[ResponseFormat] seed : typing.Optional[int] tools : typing.Optional[typing.Sequence[Tool]] tool_choice : typing.Optional[typing.Dict[str, typing.Optional[typing.Any]]] parallel_tool_calls : typing.Optional[bool] store : typing.Optional[bool] metadata : typing.Optional[typing.Dict[str, str]] reasoning_effort : typing.Optional[str] service_tier : typing.Optional[str] functions : typing.Optional[typing.Sequence[Function]] function_call : typing.Optional[FunctionCall] request_options : typing.Optional[RequestOptions] Request-specific configuration. Returns ------- HttpResponse[ChatCompletionResponse] Chat completions response """ _response = self._client_wrapper.httpx_client.request( "v1/private/chat/completions", method="POST", json={ "model": model, "messages": messages, "temperature": temperature, "top_p": top_p, "n": n, "stream": stream, "stream_options": convert_and_respect_annotation_metadata( object_=stream_options, annotation=StreamOptions, direction="write" ), "stop": stop, "max_tokens": max_tokens, "max_completion_tokens": max_completion_tokens, "presence_penalty": presence_penalty, "frequency_penalty": frequency_penalty, "logit_bias": logit_bias, "user": user, "response_format": convert_and_respect_annotation_metadata( object_=response_format, annotation=ResponseFormat, direction="write" ), "seed": seed, "tools": convert_and_respect_annotation_metadata( object_=tools, annotation=typing.Sequence[Tool], direction="write" ), "tool_choice": tool_choice, "parallel_tool_calls": parallel_tool_calls, "store": store, "metadata": metadata, "reasoning_effort": reasoning_effort, "service_tier": service_tier, "functions": convert_and_respect_annotation_metadata( object_=functions, annotation=typing.Sequence[Function], direction="write" ), "function_call": convert_and_respect_annotation_metadata( object_=function_call, annotation=FunctionCall, direction="write" ), }, headers={ "content-type": "application/json", }, request_options=request_options, omit=OMIT, ) try: if 200 <= _response.status_code < 300: _data = typing.cast( ChatCompletionResponse, parse_obj_as( type_=ChatCompletionResponse, # type: ignore object_=_response.json(), ), ) return HttpResponse(response=_response, data=_data) _response_json = _response.json() except JSONDecodeError: raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text) raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json) class AsyncRawChatCompletionsClient: def __init__(self, *, client_wrapper: AsyncClientWrapper): self._client_wrapper = client_wrapper async def create_chat_completions( self, *, model: typing.Optional[str] = OMIT, messages: typing.Optional[typing.Sequence[Message]] = OMIT, temperature: typing.Optional[float] = OMIT, top_p: typing.Optional[float] = OMIT, n: typing.Optional[int] = OMIT, stream: typing.Optional[bool] = OMIT, stream_options: typing.Optional[StreamOptions] = OMIT, stop: typing.Optional[typing.Sequence[str]] = OMIT, max_tokens: typing.Optional[int] = OMIT, max_completion_tokens: typing.Optional[int] = OMIT, presence_penalty: typing.Optional[float] = OMIT, frequency_penalty: typing.Optional[float] = OMIT, logit_bias: typing.Optional[typing.Dict[str, int]] = OMIT, user: typing.Optional[str] = OMIT, response_format: typing.Optional[ResponseFormat] = OMIT, seed: typing.Optional[int] = OMIT, tools: typing.Optional[typing.Sequence[Tool]] = OMIT, tool_choice: typing.Optional[typing.Dict[str, typing.Optional[typing.Any]]] = OMIT, parallel_tool_calls: typing.Optional[bool] = OMIT, store: typing.Optional[bool] = OMIT, metadata: typing.Optional[typing.Dict[str, str]] = OMIT, reasoning_effort: typing.Optional[str] = OMIT, service_tier: typing.Optional[str] = OMIT, functions: typing.Optional[typing.Sequence[Function]] = OMIT, function_call: typing.Optional[FunctionCall] = OMIT, request_options: typing.Optional[RequestOptions] = None, ) -> AsyncHttpResponse[ChatCompletionResponse]: """ Create chat completions Parameters ---------- model : typing.Optional[str] messages : typing.Optional[typing.Sequence[Message]] temperature : typing.Optional[float] top_p : typing.Optional[float] n : typing.Optional[int] stream : typing.Optional[bool] stream_options : typing.Optional[StreamOptions] stop : typing.Optional[typing.Sequence[str]] max_tokens : typing.Optional[int] max_completion_tokens : typing.Optional[int] presence_penalty : typing.Optional[float] frequency_penalty : typing.Optional[float] logit_bias : typing.Optional[typing.Dict[str, int]] user : typing.Optional[str] response_format : typing.Optional[ResponseFormat] seed : typing.Optional[int] tools : typing.Optional[typing.Sequence[Tool]] tool_choice : typing.Optional[typing.Dict[str, typing.Optional[typing.Any]]] parallel_tool_calls : typing.Optional[bool] store : typing.Optional[bool] metadata : typing.Optional[typing.Dict[str, str]] reasoning_effort : typing.Optional[str] service_tier : typing.Optional[str] functions : typing.Optional[typing.Sequence[Function]] function_call : typing.Optional[FunctionCall] request_options : typing.Optional[RequestOptions] Request-specific configuration. Returns ------- AsyncHttpResponse[ChatCompletionResponse] Chat completions response """ _response = await self._client_wrapper.httpx_client.request( "v1/private/chat/completions", method="POST", json={ "model": model, "messages": messages, "temperature": temperature, "top_p": top_p, "n": n, "stream": stream, "stream_options": convert_and_respect_annotation_metadata( object_=stream_options, annotation=StreamOptions, direction="write" ), "stop": stop, "max_tokens": max_tokens, "max_completion_tokens": max_completion_tokens, "presence_penalty": presence_penalty, "frequency_penalty": frequency_penalty, "logit_bias": logit_bias, "user": user, "response_format": convert_and_respect_annotation_metadata( object_=response_format, annotation=ResponseFormat, direction="write" ), "seed": seed, "tools": convert_and_respect_annotation_metadata( object_=tools, annotation=typing.Sequence[Tool], direction="write" ), "tool_choice": tool_choice, "parallel_tool_calls": parallel_tool_calls, "store": store, "metadata": metadata, "reasoning_effort": reasoning_effort, "service_tier": service_tier, "functions": convert_and_respect_annotation_metadata( object_=functions, annotation=typing.Sequence[Function], direction="write" ), "function_call": convert_and_respect_annotation_metadata( object_=function_call, annotation=FunctionCall, direction="write" ), }, headers={ "content-type": "application/json", }, request_options=request_options, omit=OMIT, ) try: if 200 <= _response.status_code < 300: _data = typing.cast( ChatCompletionResponse, parse_obj_as( type_=ChatCompletionResponse, # type: ignore object_=_response.json(), ), ) return AsyncHttpResponse(response=_response, data=_data) _response_json = _response.json() except JSONDecodeError: raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text) raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)