vellum-ai 1.1.2__py3-none-any.whl → 1.1.4__py3-none-any.whl

vellum/client/__init__.py

@@ -1,90 +1,62 @@
 # This file was auto-generated by Fern from our API Definition.
 
-import typing
-from .environment import VellumEnvironment
-from .types.api_version_enum import ApiVersionEnum
 import os
+import typing
+
 import httpx
-from .core.client_wrapper import SyncClientWrapper
-from .raw_client import RawVellum
-from .resources.ad_hoc.client import AdHocClient
-from .resources.container_images.client import ContainerImagesClient
-from .resources.deployments.client import DeploymentsClient
-from .resources.document_indexes.client import DocumentIndexesClient
-from .resources.documents.client import DocumentsClient
-from .resources.folder_entities.client import FolderEntitiesClient
-from .resources.metric_definitions.client import MetricDefinitionsClient
-from .resources.ml_models.client import MlModelsClient
-from .resources.organizations.client import OrganizationsClient
-from .resources.prompts.client import PromptsClient
-from .resources.sandboxes.client import SandboxesClient
-from .resources.test_suite_runs.client import TestSuiteRunsClient
-from .resources.test_suites.client import TestSuitesClient
-from .resources.workflow_deployments.client import WorkflowDeploymentsClient
-from .resources.workflow_executions.client import WorkflowExecutionsClient
-from .resources.workflow_sandboxes.client import WorkflowSandboxesClient
-from .resources.workflows.client import WorkflowsClient
-from .resources.workspace_secrets.client import WorkspaceSecretsClient
-from .resources.workspaces.client import WorkspacesClient
-from .types.method_enum import MethodEnum
-from .types.execute_api_request_body import ExecuteApiRequestBody
-from .types.execute_api_request_headers_value import ExecuteApiRequestHeadersValue
-from .types.execute_api_request_bearer_token import ExecuteApiRequestBearerToken
+from .core.api_error import ApiError
+from .core.client_wrapper import AsyncClientWrapper, SyncClientWrapper
 from .core.request_options import RequestOptions
-from .types.execute_api_response import ExecuteApiResponse
+from .environment import VellumEnvironment
+from .raw_client import AsyncRawVellum, RawVellum
+from .resources.ad_hoc.client import AdHocClient, AsyncAdHocClient
+from .resources.container_images.client import AsyncContainerImagesClient, ContainerImagesClient
+from .resources.deployments.client import AsyncDeploymentsClient, DeploymentsClient
+from .resources.document_indexes.client import AsyncDocumentIndexesClient, DocumentIndexesClient
+from .resources.documents.client import AsyncDocumentsClient, DocumentsClient
+from .resources.folder_entities.client import AsyncFolderEntitiesClient, FolderEntitiesClient
+from .resources.metric_definitions.client import AsyncMetricDefinitionsClient, MetricDefinitionsClient
+from .resources.ml_models.client import AsyncMlModelsClient, MlModelsClient
+from .resources.organizations.client import AsyncOrganizationsClient, OrganizationsClient
+from .resources.prompts.client import AsyncPromptsClient, PromptsClient
+from .resources.sandboxes.client import AsyncSandboxesClient, SandboxesClient
+from .resources.test_suite_runs.client import AsyncTestSuiteRunsClient, TestSuiteRunsClient
+from .resources.test_suites.client import AsyncTestSuitesClient, TestSuitesClient
+from .resources.workflow_deployments.client import AsyncWorkflowDeploymentsClient, WorkflowDeploymentsClient
+from .resources.workflow_executions.client import AsyncWorkflowExecutionsClient, WorkflowExecutionsClient
+from .resources.workflow_sandboxes.client import AsyncWorkflowSandboxesClient, WorkflowSandboxesClient
+from .resources.workflows.client import AsyncWorkflowsClient, WorkflowsClient
+from .resources.workspace_secrets.client import AsyncWorkspaceSecretsClient, WorkspaceSecretsClient
+from .resources.workspaces.client import AsyncWorkspacesClient, WorkspacesClient
+from .types.api_version_enum import ApiVersionEnum
+from .types.code_execution_package import CodeExecutionPackage
 from .types.code_execution_runtime import CodeExecutionRuntime
 from .types.code_executor_input import CodeExecutorInput
-from .types.code_execution_package import CodeExecutionPackage
-from .types.vellum_variable_type import VellumVariableType
 from .types.code_executor_response import CodeExecutorResponse
-from .types.prompt_deployment_input_request import PromptDeploymentInputRequest
-from .types.prompt_deployment_expand_meta_request import PromptDeploymentExpandMetaRequest
-from .types.raw_prompt_execution_overrides_request import RawPromptExecutionOverridesRequest
-from .types.execute_prompt_response import ExecutePromptResponse
+from .types.execute_api_request_bearer_token import ExecuteApiRequestBearerToken
+from .types.execute_api_request_body import ExecuteApiRequestBody
+from .types.execute_api_request_headers_value import ExecuteApiRequestHeadersValue
+from .types.execute_api_response import ExecuteApiResponse
 from .types.execute_prompt_event import ExecutePromptEvent
-from .core.serialization import convert_and_respect_annotation_metadata
-from .core.pydantic_utilities import parse_obj_as
-import json
-from .errors.bad_request_error import BadRequestError
-from .errors.forbidden_error import ForbiddenError
-from .errors.not_found_error import NotFoundError
-from .errors.internal_server_error import InternalServerError
-from json.decoder import JSONDecodeError
-from .core.api_error import ApiError
-from .types.workflow_request_input_request import WorkflowRequestInputRequest
-from .types.workflow_expand_meta_request import WorkflowExpandMetaRequest
+from .types.execute_prompt_response import ExecutePromptResponse
 from .types.execute_workflow_response import ExecuteWorkflowResponse
-from .types.workflow_execution_event_type import WorkflowExecutionEventType
-from .types.workflow_stream_event import WorkflowStreamEvent
-from .types.generate_request import GenerateRequest
 from .types.generate_options_request import GenerateOptionsRequest
+from .types.generate_request import GenerateRequest
 from .types.generate_response import GenerateResponse
 from .types.generate_stream_response import GenerateStreamResponse
+from .types.method_enum import MethodEnum
+from .types.prompt_deployment_expand_meta_request import PromptDeploymentExpandMetaRequest
+from .types.prompt_deployment_input_request import PromptDeploymentInputRequest
+from .types.raw_prompt_execution_overrides_request import RawPromptExecutionOverridesRequest
 from .types.search_request_options_request import SearchRequestOptionsRequest
 from .types.search_response import SearchResponse
 from .types.submit_completion_actual_request import SubmitCompletionActualRequest
 from .types.submit_workflow_execution_actual_request import SubmitWorkflowExecutionActualRequest
-from .core.client_wrapper import AsyncClientWrapper
-from .raw_client import AsyncRawVellum
-from .resources.ad_hoc.client import AsyncAdHocClient
-from .resources.container_images.client import AsyncContainerImagesClient
-from .resources.deployments.client import AsyncDeploymentsClient
-from .resources.document_indexes.client import AsyncDocumentIndexesClient
-from .resources.documents.client import AsyncDocumentsClient
-from .resources.folder_entities.client import AsyncFolderEntitiesClient
-from .resources.metric_definitions.client import AsyncMetricDefinitionsClient
-from .resources.ml_models.client import AsyncMlModelsClient
-from .resources.organizations.client import AsyncOrganizationsClient
-from .resources.prompts.client import AsyncPromptsClient
-from .resources.sandboxes.client import AsyncSandboxesClient
-from .resources.test_suite_runs.client import AsyncTestSuiteRunsClient
-from .resources.test_suites.client import AsyncTestSuitesClient
-from .resources.workflow_deployments.client import AsyncWorkflowDeploymentsClient
-from .resources.workflow_executions.client import AsyncWorkflowExecutionsClient
-from .resources.workflow_sandboxes.client import AsyncWorkflowSandboxesClient
-from .resources.workflows.client import AsyncWorkflowsClient
-from .resources.workspace_secrets.client import AsyncWorkspaceSecretsClient
-from .resources.workspaces.client import AsyncWorkspacesClient
+from .types.vellum_variable_type import VellumVariableType
+from .types.workflow_execution_event_type import WorkflowExecutionEventType
+from .types.workflow_expand_meta_request import WorkflowExpandMetaRequest
+from .types.workflow_request_input_request import WorkflowRequestInputRequest
+from .types.workflow_stream_event import WorkflowStreamEvent
 
 # this is used as the default value for optional parameters
 OMIT = typing.cast(typing.Any, ...)
@@ -107,6 +79,9 @@ class Vellum:
 
     api_version : typing.Optional[ApiVersionEnum]
     api_key : str
+    headers : typing.Optional[typing.Dict[str, str]]
+        Additional headers to send with every request.
+
     timeout : typing.Optional[float]
         The timeout to be used, in seconds, for requests. By default there is no timeout set, unless a custom httpx client is used, in which case this default is not enforced.
 
@@ -132,6 +107,7 @@ class Vellum:
         environment: VellumEnvironment = VellumEnvironment.PRODUCTION,
         api_version: typing.Optional[ApiVersionEnum] = os.getenv("VELLUM_API_VERSION", "2025-07-30"),
         api_key: str,
+        headers: typing.Optional[typing.Dict[str, str]] = None,
         timeout: typing.Optional[float] = None,
         follow_redirects: typing.Optional[bool] = True,
         httpx_client: typing.Optional[httpx.Client] = None,
@@ -143,6 +119,7 @@ class Vellum:
             environment=environment,
             api_version=api_version,
             api_key=api_key,
+            headers=headers,
             httpx_client=httpx_client
             if httpx_client is not None
             else httpx.Client(timeout=_defaulted_timeout, follow_redirects=follow_redirects)
@@ -225,7 +202,7 @@ class Vellum:
             url="x",
         )
         """
-        response = self._raw_client.execute_api(
+        _response = self._raw_client.execute_api(
             url=url,
             method=method,
             body=body,
@@ -233,7 +210,7 @@ class Vellum:
             bearer_token=bearer_token,
             request_options=request_options,
         )
-        return response.data
+        return _response.data
 
     def execute_code(
         self,
@@ -300,7 +277,7 @@ class Vellum:
             output_type="STRING",
         )
         """
-        response = self._raw_client.execute_code(
+        _response = self._raw_client.execute_code(
             code=code,
             runtime=runtime,
             input_values=input_values,
@@ -308,7 +285,7 @@ class Vellum:
             output_type=output_type,
             request_options=request_options,
         )
-        return response.data
+        return _response.data
 
     def execute_prompt(
         self,
@@ -385,7 +362,7 @@ class Vellum:
             ],
         )
         """
-        response = self._raw_client.execute_prompt(
+        _response = self._raw_client.execute_prompt(
             inputs=inputs,
             prompt_deployment_id=prompt_deployment_id,
             prompt_deployment_name=prompt_deployment_name,
@@ -397,7 +374,7 @@ class Vellum:
             metadata=metadata,
             request_options=request_options,
         )
-        return response.data
+        return _response.data
 
     def execute_prompt_stream(
         self,
@@ -476,98 +453,19 @@ class Vellum:
         for chunk in response:
             yield chunk
         """
-        with self._raw_client._client_wrapper.httpx_client.stream(
-            "v1/execute-prompt-stream",
-            base_url=self._raw_client._client_wrapper.get_environment().predict,
-            method="POST",
-            json={
-                "inputs": convert_and_respect_annotation_metadata(
-                    object_=inputs, annotation=typing.Sequence[PromptDeploymentInputRequest], direction="write"
-                ),
-                "prompt_deployment_id": prompt_deployment_id,
-                "prompt_deployment_name": prompt_deployment_name,
-                "release_tag": release_tag,
-                "external_id": external_id,
-                "expand_meta": convert_and_respect_annotation_metadata(
-                    object_=expand_meta,
-                    annotation=typing.Optional[PromptDeploymentExpandMetaRequest],
-                    direction="write",
-                ),
-                "raw_overrides": convert_and_respect_annotation_metadata(
-                    object_=raw_overrides,
-                    annotation=typing.Optional[RawPromptExecutionOverridesRequest],
-                    direction="write",
-                ),
-                "expand_raw": expand_raw,
-                "metadata": metadata,
-            },
-            headers={
-                "content-type": "application/json",
-            },
+        with self._raw_client.execute_prompt_stream(
+            inputs=inputs,
+            prompt_deployment_id=prompt_deployment_id,
+            prompt_deployment_name=prompt_deployment_name,
+            release_tag=release_tag,
+            external_id=external_id,
+            expand_meta=expand_meta,
+            raw_overrides=raw_overrides,
+            expand_raw=expand_raw,
+            metadata=metadata,
             request_options=request_options,
-            omit=OMIT,
-        ) as _response:
-            try:
-                if 200 <= _response.status_code < 300:
-                    for _text in _response.iter_lines():
-                        try:
-                            if len(_text) == 0:
-                                continue
-                            yield typing.cast(
-                                ExecutePromptEvent,
-                                parse_obj_as(
-                                    type_=ExecutePromptEvent,  # type: ignore
-                                    object_=json.loads(_text),
-                                ),
-                            )
-                        except Exception:
-                            pass
-                    return
-                _response.read()
-                if _response.status_code == 400:
-                    raise BadRequestError(
-                        typing.cast(
-                            typing.Optional[typing.Any],
-                            parse_obj_as(
-                                type_=typing.Optional[typing.Any],  # type: ignore
-                                object_=_response.json(),
-                            ),
-                        )
-                    )
-                if _response.status_code == 403:
-                    raise ForbiddenError(
-                        typing.cast(
-                            typing.Optional[typing.Any],
-                            parse_obj_as(
-                                type_=typing.Optional[typing.Any],  # type: ignore
-                                object_=_response.json(),
-                            ),
-                        )
-                    )
-                if _response.status_code == 404:
-                    raise NotFoundError(
-                        typing.cast(
-                            typing.Optional[typing.Any],
-                            parse_obj_as(
-                                type_=typing.Optional[typing.Any],  # type: ignore
-                                object_=_response.json(),
-                            ),
-                        )
-                    )
-                if _response.status_code == 500:
-                    raise InternalServerError(
-                        typing.cast(
-                            typing.Optional[typing.Any],
-                            parse_obj_as(
-                                type_=typing.Optional[typing.Any],  # type: ignore
-                                object_=_response.json(),
-                            ),
-                        )
-                    )
-                _response_json = _response.json()
-            except JSONDecodeError:
-                raise ApiError(status_code=_response.status_code, body=_response.text)
-            raise ApiError(status_code=_response.status_code, body=_response_json)
+        ) as r:
+            yield from r.data
 
     def execute_workflow(
         self,
@@ -636,7 +534,7 @@ class Vellum:
             ],
         )
         """
-        response = self._raw_client.execute_workflow(
+        _response = self._raw_client.execute_workflow(
             inputs=inputs,
             expand_meta=expand_meta,
             workflow_deployment_id=workflow_deployment_id,
@@ -646,7 +544,7 @@ class Vellum:
             metadata=metadata,
             request_options=request_options,
         )
-        return response.data
+        return _response.data
 
     def execute_workflow_stream(
         self,
@@ -721,81 +619,18 @@ class Vellum:
         for chunk in response:
             yield chunk
         """
-        with self._raw_client._client_wrapper.httpx_client.stream(
-            "v1/execute-workflow-stream",
-            base_url=self._raw_client._client_wrapper.get_environment().predict,
-            method="POST",
-            json={
-                "inputs": convert_and_respect_annotation_metadata(
-                    object_=inputs, annotation=typing.Sequence[WorkflowRequestInputRequest], direction="write"
-                ),
-                "expand_meta": convert_and_respect_annotation_metadata(
-                    object_=expand_meta, annotation=typing.Optional[WorkflowExpandMetaRequest], direction="write"
-                ),
-                "workflow_deployment_id": workflow_deployment_id,
-                "workflow_deployment_name": workflow_deployment_name,
-                "release_tag": release_tag,
-                "external_id": external_id,
-                "event_types": event_types,
-                "metadata": metadata,
-            },
-            headers={
-                "content-type": "application/json",
-            },
+        with self._raw_client.execute_workflow_stream(
+            inputs=inputs,
+            expand_meta=expand_meta,
+            workflow_deployment_id=workflow_deployment_id,
+            workflow_deployment_name=workflow_deployment_name,
+            release_tag=release_tag,
+            external_id=external_id,
+            event_types=event_types,
+            metadata=metadata,
             request_options=request_options,
-            omit=OMIT,
-        ) as _response:
-            try:
-                if 200 <= _response.status_code < 300:
-                    for _text in _response.iter_lines():
-                        try:
-                            if len(_text) == 0:
-                                continue
-                            yield typing.cast(
-                                WorkflowStreamEvent,
-                                parse_obj_as(
-                                    type_=WorkflowStreamEvent,  # type: ignore
-                                    object_=json.loads(_text),
-                                ),
-                            )
-                        except Exception:
-                            pass
-                    return
-                _response.read()
-                if _response.status_code == 400:
-                    raise BadRequestError(
-                        typing.cast(
-                            typing.Optional[typing.Any],
-                            parse_obj_as(
-                                type_=typing.Optional[typing.Any],  # type: ignore
-                                object_=_response.json(),
-                            ),
-                        )
-                    )
-                if _response.status_code == 404:
-                    raise NotFoundError(
-                        typing.cast(
-                            typing.Optional[typing.Any],
-                            parse_obj_as(
-                                type_=typing.Optional[typing.Any],  # type: ignore
-                                object_=_response.json(),
-                            ),
-                        )
-                    )
-                if _response.status_code == 500:
-                    raise InternalServerError(
-                        typing.cast(
-                            typing.Optional[typing.Any],
-                            parse_obj_as(
-                                type_=typing.Optional[typing.Any],  # type: ignore
-                                object_=_response.json(),
-                            ),
-                        )
-                    )
-                _response_json = _response.json()
-            except JSONDecodeError:
-                raise ApiError(status_code=_response.status_code, body=_response.text)
-            raise ApiError(status_code=_response.status_code, body=_response_json)
+        ) as r:
+            yield from r.data
 
     def generate(
         self,
@@ -853,14 +688,14 @@ class Vellum:
             ],
         )
         """
-        response = self._raw_client.generate(
+        _response = self._raw_client.generate(
             requests=requests,
             deployment_id=deployment_id,
             deployment_name=deployment_name,
             options=options,
             request_options=request_options,
         )
-        return response.data
+        return _response.data
 
     def generate_stream(
         self,
@@ -920,87 +755,14 @@ class Vellum:
         for chunk in response:
             yield chunk
         """
-        with self._raw_client._client_wrapper.httpx_client.stream(
-            "v1/generate-stream",
-            base_url=self._raw_client._client_wrapper.get_environment().predict,
-            method="POST",
-            json={
-                "deployment_id": deployment_id,
-                "deployment_name": deployment_name,
-                "requests": convert_and_respect_annotation_metadata(
-                    object_=requests, annotation=typing.Sequence[GenerateRequest], direction="write"
-                ),
-                "options": convert_and_respect_annotation_metadata(
-                    object_=options, annotation=typing.Optional[GenerateOptionsRequest], direction="write"
-                ),
-            },
-            headers={
-                "content-type": "application/json",
-            },
+        with self._raw_client.generate_stream(
+            requests=requests,
+            deployment_id=deployment_id,
+            deployment_name=deployment_name,
+            options=options,
             request_options=request_options,
-            omit=OMIT,
-        ) as _response:
-            try:
-                if 200 <= _response.status_code < 300:
-                    for _text in _response.iter_lines():
-                        try:
-                            if len(_text) == 0:
-                                continue
-                            yield typing.cast(
-                                GenerateStreamResponse,
-                                parse_obj_as(
-                                    type_=GenerateStreamResponse,  # type: ignore
-                                    object_=json.loads(_text),
-                                ),
-                            )
-                        except Exception:
-                            pass
-                    return
-                _response.read()
-                if _response.status_code == 400:
-                    raise BadRequestError(
-                        typing.cast(
-                            typing.Optional[typing.Any],
-                            parse_obj_as(
-                                type_=typing.Optional[typing.Any],  # type: ignore
-                                object_=_response.json(),
-                            ),
-                        )
-                    )
-                if _response.status_code == 403:
-                    raise ForbiddenError(
-                        typing.cast(
-                            typing.Optional[typing.Any],
-                            parse_obj_as(
-                                type_=typing.Optional[typing.Any],  # type: ignore
-                                object_=_response.json(),
-                            ),
-                        )
-                    )
-                if _response.status_code == 404:
-                    raise NotFoundError(
-                        typing.cast(
-                            typing.Optional[typing.Any],
-                            parse_obj_as(
-                                type_=typing.Optional[typing.Any],  # type: ignore
-                                object_=_response.json(),
-                            ),
-                        )
-                    )
-                if _response.status_code == 500:
-                    raise InternalServerError(
-                        typing.cast(
-                            typing.Optional[typing.Any],
-                            parse_obj_as(
-                                type_=typing.Optional[typing.Any],  # type: ignore
-                                object_=_response.json(),
-                            ),
-                        )
-                    )
-                _response_json = _response.json()
-            except JSONDecodeError:
-                raise ApiError(status_code=_response.status_code, body=_response.text)
-            raise ApiError(status_code=_response.status_code, body=_response_json)
+        ) as r:
+            yield from r.data
 
     def search(
         self,
@@ -1052,7 +814,7 @@ class Vellum:
             query="x",
         )
         """
-        response = self._raw_client.search(
+        _response = self._raw_client.search(
             query=query,
             index_id=index_id,
             index_name=index_name,
@@ -1060,7 +822,7 @@ class Vellum:
             document_index=document_index,
             request_options=request_options,
         )
-        return response.data
+        return _response.data
 
     def submit_completion_actuals(
         self,
@@ -1103,13 +865,13 @@ class Vellum:
             actuals=[SubmitCompletionActualRequest(), SubmitCompletionActualRequest()],
         )
         """
-        response = self._raw_client.submit_completion_actuals(
+        _response = self._raw_client.submit_completion_actuals(
             actuals=actuals,
             deployment_id=deployment_id,
             deployment_name=deployment_name,
             request_options=request_options,
         )
-        return response.data
+        return _response.data
 
     def submit_workflow_execution_actuals(
         self,
@@ -1157,13 +919,10 @@ class Vellum:
             ],
         )
         """
-        response = self._raw_client.submit_workflow_execution_actuals(
-            actuals=actuals,
-            execution_id=execution_id,
-            external_id=external_id,
-            request_options=request_options,
+        _response = self._raw_client.submit_workflow_execution_actuals(
+            actuals=actuals, execution_id=execution_id, external_id=external_id, request_options=request_options
         )
-        return response.data
+        return _response.data
 
 
 class AsyncVellum:
@@ -1183,6 +942,9 @@ class AsyncVellum:
 
     api_version : typing.Optional[ApiVersionEnum]
     api_key : str
+    headers : typing.Optional[typing.Dict[str, str]]
+        Additional headers to send with every request.
+
     timeout : typing.Optional[float]
         The timeout to be used, in seconds, for requests. By default there is no timeout set, unless a custom httpx client is used, in which case this default is not enforced.
 
@@ -1208,6 +970,7 @@ class AsyncVellum:
         environment: VellumEnvironment = VellumEnvironment.PRODUCTION,
         api_version: typing.Optional[ApiVersionEnum] = os.getenv("VELLUM_API_VERSION", "2025-07-30"),
         api_key: str,
+        headers: typing.Optional[typing.Dict[str, str]] = None,
         timeout: typing.Optional[float] = None,
         follow_redirects: typing.Optional[bool] = True,
         httpx_client: typing.Optional[httpx.AsyncClient] = None,
@@ -1219,6 +982,7 @@ class AsyncVellum:
             environment=environment,
             api_version=api_version,
             api_key=api_key,
+            headers=headers,
             httpx_client=httpx_client
             if httpx_client is not None
             else httpx.AsyncClient(timeout=_defaulted_timeout, follow_redirects=follow_redirects)
@@ -1309,7 +1073,7 @@ class AsyncVellum:
 
         asyncio.run(main())
         """
-        response = await self._raw_client.execute_api(
+        _response = await self._raw_client.execute_api(
             url=url,
             method=method,
             body=body,
@@ -1317,7 +1081,7 @@ class AsyncVellum:
             bearer_token=bearer_token,
             request_options=request_options,
         )
-        return response.data
+        return _response.data
 
     async def execute_code(
         self,
@@ -1392,7 +1156,7 @@ class AsyncVellum:
 
         asyncio.run(main())
         """
-        response = await self._raw_client.execute_code(
+        _response = await self._raw_client.execute_code(
             code=code,
             runtime=runtime,
             input_values=input_values,
@@ -1400,7 +1164,7 @@ class AsyncVellum:
             output_type=output_type,
             request_options=request_options,
         )
-        return response.data
+        return _response.data
 
     async def execute_prompt(
         self,
@@ -1485,7 +1249,7 @@ class AsyncVellum:
 
         asyncio.run(main())
         """
-        response = await self._raw_client.execute_prompt(
+        _response = await self._raw_client.execute_prompt(
             inputs=inputs,
             prompt_deployment_id=prompt_deployment_id,
             prompt_deployment_name=prompt_deployment_name,
@@ -1497,7 +1261,7 @@ class AsyncVellum:
             metadata=metadata,
             request_options=request_options,
         )
-        return response.data
+        return _response.data
 
     async def execute_prompt_stream(
         self,
@@ -1584,98 +1348,20 @@ class AsyncVellum:
 
         asyncio.run(main())
         """
-        async with self._raw_client._client_wrapper.httpx_client.stream(
-            "v1/execute-prompt-stream",
-            base_url=self._raw_client._client_wrapper.get_environment().predict,
-            method="POST",
-            json={
-                "inputs": convert_and_respect_annotation_metadata(
-                    object_=inputs, annotation=typing.Sequence[PromptDeploymentInputRequest], direction="write"
-                ),
-                "prompt_deployment_id": prompt_deployment_id,
-                "prompt_deployment_name": prompt_deployment_name,
-                "release_tag": release_tag,
-                "external_id": external_id,
-                "expand_meta": convert_and_respect_annotation_metadata(
-                    object_=expand_meta,
-                    annotation=typing.Optional[PromptDeploymentExpandMetaRequest],
-                    direction="write",
-                ),
-                "raw_overrides": convert_and_respect_annotation_metadata(
-                    object_=raw_overrides,
-                    annotation=typing.Optional[RawPromptExecutionOverridesRequest],
-                    direction="write",
-                ),
-                "expand_raw": expand_raw,
-                "metadata": metadata,
-            },
-            headers={
-                "content-type": "application/json",
-            },
+        async with self._raw_client.execute_prompt_stream(
+            inputs=inputs,
+            prompt_deployment_id=prompt_deployment_id,
+            prompt_deployment_name=prompt_deployment_name,
+            release_tag=release_tag,
+            external_id=external_id,
+            expand_meta=expand_meta,
+            raw_overrides=raw_overrides,
+            expand_raw=expand_raw,
+            metadata=metadata,
             request_options=request_options,
-            omit=OMIT,
-        ) as _response:
-            try:
-                if 200 <= _response.status_code < 300:
-                    async for _text in _response.aiter_lines():
-                        try:
-                            if len(_text) == 0:
-                                continue
-                            yield typing.cast(
-                                ExecutePromptEvent,
-                                parse_obj_as(
-                                    type_=ExecutePromptEvent,  # type: ignore
-                                    object_=json.loads(_text),
-                                ),
-                            )
-                        except Exception:
-                            pass
-                    return
-                await _response.aread()
-                if _response.status_code == 400:
-                    raise BadRequestError(
-                        typing.cast(
-                            typing.Optional[typing.Any],
-                            parse_obj_as(
-                                type_=typing.Optional[typing.Any],  # type: ignore
-                                object_=_response.json(),
-                            ),
-                        )
-                    )
-                if _response.status_code == 403:
-                    raise ForbiddenError(
-                        typing.cast(
-                            typing.Optional[typing.Any],
-                            parse_obj_as(
-                                type_=typing.Optional[typing.Any],  # type: ignore
-                                object_=_response.json(),
-                            ),
-                        )
-                    )
-                if _response.status_code == 404:
-                    raise NotFoundError(
-                        typing.cast(
-                            typing.Optional[typing.Any],
-                            parse_obj_as(
-                                type_=typing.Optional[typing.Any],  # type: ignore
-                                object_=_response.json(),
-                            ),
-                        )
-                    )
-                if _response.status_code == 500:
-                    raise InternalServerError(
-                        typing.cast(
-                            typing.Optional[typing.Any],
-                            parse_obj_as(
-                                type_=typing.Optional[typing.Any],  # type: ignore
-                                object_=_response.json(),
-                            ),
-                        )
-                    )
-                _response_json = _response.json()
-            except JSONDecodeError:
-                raise ApiError(status_code=_response.status_code, body=_response.text)
-            raise ApiError(status_code=_response.status_code, body=_response_json)
+        ) as r:
+            async for _chunk in r.data:
+                yield _chunk
 
     async def execute_workflow(
         self,
@@ -1752,7 +1438,7 @@ class AsyncVellum:
 
         asyncio.run(main())
         """
-        response = await self._raw_client.execute_workflow(
+        _response = await self._raw_client.execute_workflow(
             inputs=inputs,
             expand_meta=expand_meta,
             workflow_deployment_id=workflow_deployment_id,
@@ -1762,7 +1448,7 @@ class AsyncVellum:
             metadata=metadata,
             request_options=request_options,
         )
-        return response.data
+        return _response.data
 
     async def execute_workflow_stream(
         self,
@@ -1845,81 +1531,19 @@ class AsyncVellum:
 
         asyncio.run(main())
         """
-        async with self._raw_client._client_wrapper.httpx_client.stream(
-            "v1/execute-workflow-stream",
-            base_url=self._raw_client._client_wrapper.get_environment().predict,
-            method="POST",
-            json={
-                "inputs": convert_and_respect_annotation_metadata(
-                    object_=inputs, annotation=typing.Sequence[WorkflowRequestInputRequest], direction="write"
-                ),
-                "expand_meta": convert_and_respect_annotation_metadata(
-                    object_=expand_meta, annotation=typing.Optional[WorkflowExpandMetaRequest], direction="write"
-                ),
-                "workflow_deployment_id": workflow_deployment_id,
-                "workflow_deployment_name": workflow_deployment_name,
-                "release_tag": release_tag,
-                "external_id": external_id,
-                "event_types": event_types,
-                "metadata": metadata,
-            },
-            headers={
-                "content-type": "application/json",
-            },
+        async with self._raw_client.execute_workflow_stream(
+            inputs=inputs,
+            expand_meta=expand_meta,
+            workflow_deployment_id=workflow_deployment_id,
+            workflow_deployment_name=workflow_deployment_name,
+            release_tag=release_tag,
+            external_id=external_id,
+            event_types=event_types,
+            metadata=metadata,
             request_options=request_options,
-            omit=OMIT,
-        ) as _response:
-            try:
-                if 200 <= _response.status_code < 300:
-                    async for _text in _response.aiter_lines():
-                        try:
-                            if len(_text) == 0:
-                                continue
-                            yield typing.cast(
-                                WorkflowStreamEvent,
-                                parse_obj_as(
-                                    type_=WorkflowStreamEvent,  # type: ignore
-                                    object_=json.loads(_text),
-                                ),
-                            )
-                        except Exception:
-                            pass
-                    return
-                await _response.aread()
-                if _response.status_code == 400:
-                    raise BadRequestError(
-                        typing.cast(
-                            typing.Optional[typing.Any],
-                            parse_obj_as(
-                                type_=typing.Optional[typing.Any],  # type: ignore
-                                object_=_response.json(),
-                            ),
-                        )
-                    )
-                if _response.status_code == 404:
-                    raise NotFoundError(
-                        typing.cast(
-                            typing.Optional[typing.Any],
-                            parse_obj_as(
-                                type_=typing.Optional[typing.Any],  # type: ignore
-                                object_=_response.json(),
-                            ),
-                        )
-                    )
-                if _response.status_code == 500:
-                    raise InternalServerError(
-                        typing.cast(
-                            typing.Optional[typing.Any],
-                            parse_obj_as(
-                                type_=typing.Optional[typing.Any],  # type: ignore
-                                object_=_response.json(),
-                            ),
-                        )
-                    )
-                _response_json = _response.json()
-            except JSONDecodeError:
-                raise ApiError(status_code=_response.status_code, body=_response.text)
-            raise ApiError(status_code=_response.status_code, body=_response_json)
+        ) as r:
+            async for _chunk in r.data:
+                yield _chunk
 
     async def generate(
         self,
@@ -1985,14 +1609,14 @@ class AsyncVellum:
 
         asyncio.run(main())
         """
-        response = await self._raw_client.generate(
+        _response = await self._raw_client.generate(
             requests=requests,
             deployment_id=deployment_id,
             deployment_name=deployment_name,
             options=options,
             request_options=request_options,
         )
-        return response.data
+        return _response.data
 
     async def generate_stream(
         self,
@@ -2060,87 +1684,15 @@ class AsyncVellum:
 
         asyncio.run(main())
         """
-        async with self._raw_client._client_wrapper.httpx_client.stream(
-            "v1/generate-stream",
-            base_url=self._raw_client._client_wrapper.get_environment().predict,
-            method="POST",
-            json={
-                "deployment_id": deployment_id,
-                "deployment_name": deployment_name,
-                "requests": convert_and_respect_annotation_metadata(
-                    object_=requests, annotation=typing.Sequence[GenerateRequest], direction="write"
-                ),
-                "options": convert_and_respect_annotation_metadata(
-                    object_=options, annotation=typing.Optional[GenerateOptionsRequest], direction="write"
-                ),
-            },
-            headers={
-                "content-type": "application/json",
-            },
+        async with self._raw_client.generate_stream(
+            requests=requests,
+            deployment_id=deployment_id,
+            deployment_name=deployment_name,
+            options=options,
             request_options=request_options,
-            omit=OMIT,
-        ) as _response:
-            try:
-                if 200 <= _response.status_code < 300:
-                    async for _text in _response.aiter_lines():
-                        try:
-                            if len(_text) == 0:
-                                continue
-                            yield typing.cast(
-                                GenerateStreamResponse,
-                                parse_obj_as(
-                                    type_=GenerateStreamResponse,  # type: ignore
-                                    object_=json.loads(_text),
-                                ),
-                            )
-                        except Exception:
-                            pass
-                    return
-                await _response.aread()
-                if _response.status_code == 400:
-                    raise BadRequestError(
-                        typing.cast(
-                            typing.Optional[typing.Any],
-                            parse_obj_as(
-                                type_=typing.Optional[typing.Any],  # type: ignore
-                                object_=_response.json(),
-                            ),
-                        )
-                    )
-                if _response.status_code == 403:
-                    raise ForbiddenError(
-                        typing.cast(
-                            typing.Optional[typing.Any],
-                            parse_obj_as(
-                                type_=typing.Optional[typing.Any],  # type: ignore
-                                object_=_response.json(),
-                            ),
-                        )
-                    )
-                if _response.status_code == 404:
-                    raise NotFoundError(
-                        typing.cast(
-                            typing.Optional[typing.Any],
-                            parse_obj_as(
-                                type_=typing.Optional[typing.Any],  # type: ignore
-                                object_=_response.json(),
-                            ),
-                        )
-                    )
-                if _response.status_code == 500:
-                    raise InternalServerError(
-                        typing.cast(
-                            typing.Optional[typing.Any],
-                            parse_obj_as(
-                                type_=typing.Optional[typing.Any],  # type: ignore
-                                object_=_response.json(),
-                            ),
-                        )
-                    )
-                _response_json = _response.json()
-            except JSONDecodeError:
-                raise ApiError(status_code=_response.status_code, body=_response.text)
-            raise ApiError(status_code=_response.status_code, body=_response_json)
+        ) as r:
+            async for _chunk in r.data:
+                yield _chunk
 
     async def search(
         self,
@@ -2200,7 +1752,7 @@ class AsyncVellum:
 
         asyncio.run(main())
         """
-        response = await self._raw_client.search(
+        _response = await self._raw_client.search(
             query=query,
             index_id=index_id,
             index_name=index_name,
@@ -2208,7 +1760,7 @@ class AsyncVellum:
             document_index=document_index,
             request_options=request_options,
         )
-        return response.data
+        return _response.data
 
     async def submit_completion_actuals(
         self,
@@ -2262,13 +1814,13 @@ class AsyncVellum:
 
         asyncio.run(main())
         """
-        response = await self._raw_client.submit_completion_actuals(
+        _response = await self._raw_client.submit_completion_actuals(
             actuals=actuals,
             deployment_id=deployment_id,
             deployment_name=deployment_name,
             request_options=request_options,
         )
-        return response.data
+        return _response.data
 
     async def submit_workflow_execution_actuals(
         self,
@@ -2324,10 +1876,7 @@ class AsyncVellum:
 
         asyncio.run(main())
         """
-        response = await self._raw_client.submit_workflow_execution_actuals(
-            actuals=actuals,
-            execution_id=execution_id,
-            external_id=external_id,
-            request_options=request_options,
+        _response = await self._raw_client.submit_workflow_execution_actuals(
+            actuals=actuals, execution_id=execution_id, external_id=external_id, request_options=request_options
         )
-        return response.data
+        return _response.data