vellum-ai 0.1.8__py3-none-any.whl → 0.1.10__py3-none-any.whl

vellum/__init__.py

@@ -35,10 +35,24 @@ from .types import (
     DocumentStatus,
     EnrichedNormalizedCompletion,
     EnvironmentEnum,
+    ErrorVariableValue,
     EvaluationParams,
     EvaluationParamsRequest,
+    ExecutePromptApiErrorResponse,
+    ExecutePromptEvent,
+    ExecutePromptEvent_Fulfilled,
+    ExecutePromptEvent_Initiated,
+    ExecutePromptEvent_Rejected,
+    ExecutePromptEvent_Streaming,
+    ExecutePromptResponse,
+    ExecutePromptResponse_Fulfilled,
+    ExecutePromptResponse_Rejected,
     ExecuteWorkflowStreamErrorResponse,
     FinishReasonEnum,
+    FulfilledEnum,
+    FulfilledExecutePromptEvent,
+    FulfilledExecutePromptResponse,
+    FulfilledPromptExecutionMeta,
     GenerateErrorResponse,
     GenerateOptionsRequest,
     GenerateRequest,
@@ -50,7 +64,11 @@ from .types import (
     GenerateStreamResult,
     GenerateStreamResultData,
     IndexingStateEnum,
+    InitiatedEnum,
+    InitiatedExecutePromptEvent,
+    InitiatedPromptExecutionMeta,
     JsonInputRequest,
+    JsonVariableValue,
     LogicalOperator,
     LogprobsEnum,
     MetadataFilterConfigRequest,
@@ -93,12 +111,18 @@ from .types import (
     PaginatedSlimDocumentList,
     ProcessingFailureReasonEnum,
     ProcessingStateEnum,
+    PromptDeploymentExpandMetaRequestRequest,
     PromptDeploymentInputRequest,
     PromptDeploymentInputRequest_ChatHistory,
     PromptDeploymentInputRequest_Json,
     PromptDeploymentInputRequest_String,
+    PromptExecutionMeta,
     PromptNodeResult,
     PromptNodeResultData,
+    PromptOutput,
+    PromptOutput_Error,
+    PromptOutput_Json,
+    PromptOutput_String,
     PromptTemplateBlock,
     PromptTemplateBlockData,
     PromptTemplateBlockDataRequest,
@@ -106,6 +130,7 @@ from .types import (
     PromptTemplateBlockPropertiesRequest,
     PromptTemplateBlockRequest,
     ProviderEnum,
+    RawPromptExecutionOverridesRequest,
     RegisterPromptErrorResponse,
     RegisterPromptModelParametersRequest,
     RegisterPromptPrompt,
@@ -116,6 +141,10 @@ from .types import (
     RegisteredPromptModelVersion,
     RegisteredPromptSandbox,
     RegisteredPromptSandboxSnapshot,
+    RejectedEnum,
+    RejectedExecutePromptEvent,
+    RejectedExecutePromptResponse,
+    RejectedPromptExecutionMeta,
     SandboxMetricInputParams,
     SandboxMetricInputParamsRequest,
     SandboxScenario,
@@ -135,7 +164,11 @@ from .types import (
     SearchResultRequest,
     SearchWeightsRequest,
     SlimDocument,
+    StreamingEnum,
+    StreamingExecutePromptEvent,
+    StreamingPromptExecutionMeta,
     StringInputRequest,
+    StringVariableValue,
     SubmitCompletionActualRequest,
     SubmitCompletionActualsErrorResponse,
     SubmitWorkflowExecutionActualRequest,
@@ -215,8 +248,10 @@ from .types import (
     WorkflowRequestInputRequest,
     WorkflowRequestInputRequest_ChatHistory,
     WorkflowRequestInputRequest_Json,
+    WorkflowRequestInputRequest_Number,
     WorkflowRequestInputRequest_String,
     WorkflowRequestJsonInputRequest,
+    WorkflowRequestNumberInputRequest,
     WorkflowRequestStringInputRequest,
     WorkflowResultEvent,
     WorkflowResultEventOutputData,
@@ -285,11 +320,25 @@ __all__ = [
     "DocumentStatus",
     "EnrichedNormalizedCompletion",
     "EnvironmentEnum",
+    "ErrorVariableValue",
     "EvaluationParams",
     "EvaluationParamsRequest",
+    "ExecutePromptApiErrorResponse",
+    "ExecutePromptEvent",
+    "ExecutePromptEvent_Fulfilled",
+    "ExecutePromptEvent_Initiated",
+    "ExecutePromptEvent_Rejected",
+    "ExecutePromptEvent_Streaming",
+    "ExecutePromptResponse",
+    "ExecutePromptResponse_Fulfilled",
+    "ExecutePromptResponse_Rejected",
     "ExecuteWorkflowStreamErrorResponse",
     "FinishReasonEnum",
     "ForbiddenError",
+    "FulfilledEnum",
+    "FulfilledExecutePromptEvent",
+    "FulfilledExecutePromptResponse",
+    "FulfilledPromptExecutionMeta",
     "GenerateErrorResponse",
     "GenerateOptionsRequest",
     "GenerateRequest",
@@ -301,8 +350,12 @@ __all__ = [
     "GenerateStreamResult",
     "GenerateStreamResultData",
     "IndexingStateEnum",
+    "InitiatedEnum",
+    "InitiatedExecutePromptEvent",
+    "InitiatedPromptExecutionMeta",
     "InternalServerError",
     "JsonInputRequest",
+    "JsonVariableValue",
     "LogicalOperator",
     "LogprobsEnum",
     "MetadataFilterConfigRequest",
@@ -346,12 +399,18 @@ __all__ = [
     "PaginatedSlimDocumentList",
     "ProcessingFailureReasonEnum",
     "ProcessingStateEnum",
+    "PromptDeploymentExpandMetaRequestRequest",
     "PromptDeploymentInputRequest",
     "PromptDeploymentInputRequest_ChatHistory",
     "PromptDeploymentInputRequest_Json",
     "PromptDeploymentInputRequest_String",
+    "PromptExecutionMeta",
     "PromptNodeResult",
     "PromptNodeResultData",
+    "PromptOutput",
+    "PromptOutput_Error",
+    "PromptOutput_Json",
+    "PromptOutput_String",
     "PromptTemplateBlock",
     "PromptTemplateBlockData",
     "PromptTemplateBlockDataRequest",
@@ -359,6 +418,7 @@ __all__ = [
     "PromptTemplateBlockPropertiesRequest",
     "PromptTemplateBlockRequest",
     "ProviderEnum",
+    "RawPromptExecutionOverridesRequest",
     "RegisterPromptErrorResponse",
     "RegisterPromptModelParametersRequest",
     "RegisterPromptPrompt",
@@ -369,6 +429,10 @@ __all__ = [
     "RegisteredPromptModelVersion",
     "RegisteredPromptSandbox",
     "RegisteredPromptSandboxSnapshot",
+    "RejectedEnum",
+    "RejectedExecutePromptEvent",
+    "RejectedExecutePromptResponse",
+    "RejectedPromptExecutionMeta",
     "SandboxMetricInputParams",
     "SandboxMetricInputParamsRequest",
     "SandboxScenario",
@@ -388,7 +452,11 @@ __all__ = [
     "SearchResultRequest",
     "SearchWeightsRequest",
     "SlimDocument",
+    "StreamingEnum",
+    "StreamingExecutePromptEvent",
+    "StreamingPromptExecutionMeta",
     "StringInputRequest",
+    "StringVariableValue",
     "SubmitCompletionActualRequest",
     "SubmitCompletionActualsErrorResponse",
     "SubmitWorkflowExecutionActualRequest",
@@ -469,8 +537,10 @@ __all__ = [
     "WorkflowRequestInputRequest",
     "WorkflowRequestInputRequest_ChatHistory",
     "WorkflowRequestInputRequest_Json",
+    "WorkflowRequestInputRequest_Number",
     "WorkflowRequestInputRequest_String",
     "WorkflowRequestJsonInputRequest",
+    "WorkflowRequestNumberInputRequest",
     "WorkflowRequestStringInputRequest",
     "WorkflowResultEvent",
     "WorkflowResultEventOutputData",

vellum/client.py

@@ -22,11 +22,15 @@ from .resources.model_versions.client import AsyncModelVersionsClient, ModelVers
 from .resources.registered_prompts.client import AsyncRegisteredPromptsClient, RegisteredPromptsClient
 from .resources.sandboxes.client import AsyncSandboxesClient, SandboxesClient
 from .resources.test_suites.client import AsyncTestSuitesClient, TestSuitesClient
-from .types.generate_error_response import GenerateErrorResponse
+from .types.execute_prompt_event import ExecutePromptEvent
+from .types.execute_prompt_response import ExecutePromptResponse
 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.prompt_deployment_expand_meta_request_request import PromptDeploymentExpandMetaRequestRequest
+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
@@ -66,6 +70,167 @@ class Vellum:
         self.sandboxes = SandboxesClient(client_wrapper=self._client_wrapper)
         self.test_suites = TestSuitesClient(client_wrapper=self._client_wrapper)
 
+    def execute_prompt(
+        self,
+        *,
+        inputs: typing.List[PromptDeploymentInputRequest],
+        prompt_deployment_id: typing.Optional[str] = OMIT,
+        prompt_deployment_name: typing.Optional[str] = OMIT,
+        release_tag: typing.Optional[str] = OMIT,
+        external_id: typing.Optional[str] = OMIT,
+        expand_meta: typing.Optional[PromptDeploymentExpandMetaRequestRequest] = OMIT,
+        raw_overrides: typing.Optional[RawPromptExecutionOverridesRequest] = OMIT,
+        expand_raw: typing.Optional[typing.List[str]] = OMIT,
+        metadata: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+    ) -> ExecutePromptResponse:
+        """
+        Executes a deployed Prompt and returns the result.
+
+        Note: This endpoint temporarily does not support prompts with function calling, support is coming soon.
+        In the meantime, we recommend still using the `/generate` endpoint for prompts with function calling.
+
+        Parameters:
+            - inputs: typing.List[PromptDeploymentInputRequest].
+
+            - prompt_deployment_id: typing.Optional[str]. The ID of the Prompt Deployment. Must provide either this or prompt_deployment_name.
+
+            - prompt_deployment_name: typing.Optional[str]. The name of the Prompt Deployment. Must provide either this or prompt_deployment_id.
+
+            - release_tag: typing.Optional[str]. Optionally specify a release tag if you want to pin to a specific release of the Prompt Deployment
+
+            - external_id: typing.Optional[str].
+
+            - expand_meta: typing.Optional[PromptDeploymentExpandMetaRequestRequest]. The name of the Prompt Deployment. Must provide either this or prompt_deployment_id.
+
+            - raw_overrides: typing.Optional[RawPromptExecutionOverridesRequest].
+
+            - expand_raw: typing.Optional[typing.List[str]]. Returns the raw API response data sent from the model host. Combined with `raw_overrides`, it can be used to access new features from models.
+
+            - metadata: typing.Optional[typing.Dict[str, typing.Any]].
+        """
+        _request: typing.Dict[str, typing.Any] = {"inputs": inputs}
+        if prompt_deployment_id is not OMIT:
+            _request["prompt_deployment_id"] = prompt_deployment_id
+        if prompt_deployment_name is not OMIT:
+            _request["prompt_deployment_name"] = prompt_deployment_name
+        if release_tag is not OMIT:
+            _request["release_tag"] = release_tag
+        if external_id is not OMIT:
+            _request["external_id"] = external_id
+        if expand_meta is not OMIT:
+            _request["expand_meta"] = expand_meta
+        if raw_overrides is not OMIT:
+            _request["raw_overrides"] = raw_overrides
+        if expand_raw is not OMIT:
+            _request["expand_raw"] = expand_raw
+        if metadata is not OMIT:
+            _request["metadata"] = metadata
+        _response = self._client_wrapper.httpx_client.request(
+            "POST",
+            urllib.parse.urljoin(f"{self._client_wrapper.get_environment().predict}/", "v1/execute-prompt"),
+            json=jsonable_encoder(_request),
+            headers=self._client_wrapper.get_headers(),
+            timeout=None,
+        )
+        if 200 <= _response.status_code < 300:
+            return pydantic.parse_obj_as(ExecutePromptResponse, _response.json())  # type: ignore
+        if _response.status_code == 400:
+            raise BadRequestError(pydantic.parse_obj_as(typing.Any, _response.json()))  # type: ignore
+        if _response.status_code == 403:
+            raise ForbiddenError(pydantic.parse_obj_as(typing.Any, _response.json()))  # type: ignore
+        if _response.status_code == 404:
+            raise NotFoundError(pydantic.parse_obj_as(typing.Any, _response.json()))  # type: ignore
+        if _response.status_code == 500:
+            raise InternalServerError(pydantic.parse_obj_as(typing.Any, _response.json()))  # type: ignore
+        try:
+            _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)
+
+    def execute_prompt_stream(
+        self,
+        *,
+        inputs: typing.List[PromptDeploymentInputRequest],
+        prompt_deployment_id: typing.Optional[str] = OMIT,
+        prompt_deployment_name: typing.Optional[str] = OMIT,
+        release_tag: typing.Optional[str] = OMIT,
+        external_id: typing.Optional[str] = OMIT,
+        expand_meta: typing.Optional[PromptDeploymentExpandMetaRequestRequest] = OMIT,
+        raw_overrides: typing.Optional[RawPromptExecutionOverridesRequest] = OMIT,
+        expand_raw: typing.Optional[typing.List[str]] = OMIT,
+        metadata: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+    ) -> typing.Iterator[ExecutePromptEvent]:
+        """
+        Executes a deployed Prompt and streams back the results.
+
+        Note: This endpoint temporarily does not support prompts with function calling, support is coming soon.
+        In the meantime, we recommend still using the `/generate-stream` endpoint for prompts with function calling
+
+        Parameters:
+            - inputs: typing.List[PromptDeploymentInputRequest].
+
+            - prompt_deployment_id: typing.Optional[str]. The ID of the Prompt Deployment. Must provide either this or prompt_deployment_name.
+
+            - prompt_deployment_name: typing.Optional[str]. The name of the Prompt Deployment. Must provide either this or prompt_deployment_id.
+
+            - release_tag: typing.Optional[str]. Optionally specify a release tag if you want to pin to a specific release of the Prompt Deployment
+
+            - external_id: typing.Optional[str].
+
+            - expand_meta: typing.Optional[PromptDeploymentExpandMetaRequestRequest]. The name of the Prompt Deployment. Must provide either this or prompt_deployment_id.
+
+            - raw_overrides: typing.Optional[RawPromptExecutionOverridesRequest].
+
+            - expand_raw: typing.Optional[typing.List[str]]. Returns the raw API response data sent from the model host. Combined with `raw_overrides`, it can be used to access new features from models.
+
+            - metadata: typing.Optional[typing.Dict[str, typing.Any]].
+        """
+        _request: typing.Dict[str, typing.Any] = {"inputs": inputs}
+        if prompt_deployment_id is not OMIT:
+            _request["prompt_deployment_id"] = prompt_deployment_id
+        if prompt_deployment_name is not OMIT:
+            _request["prompt_deployment_name"] = prompt_deployment_name
+        if release_tag is not OMIT:
+            _request["release_tag"] = release_tag
+        if external_id is not OMIT:
+            _request["external_id"] = external_id
+        if expand_meta is not OMIT:
+            _request["expand_meta"] = expand_meta
+        if raw_overrides is not OMIT:
+            _request["raw_overrides"] = raw_overrides
+        if expand_raw is not OMIT:
+            _request["expand_raw"] = expand_raw
+        if metadata is not OMIT:
+            _request["metadata"] = metadata
+        with self._client_wrapper.httpx_client.stream(
+            "POST",
+            urllib.parse.urljoin(f"{self._client_wrapper.get_environment().predict}/", "v1/execute-prompt-stream"),
+            json=jsonable_encoder(_request),
+            headers=self._client_wrapper.get_headers(),
+            timeout=None,
+        ) as _response:
+            if 200 <= _response.status_code < 300:
+                for _text in _response.iter_lines():
+                    if len(_text) == 0:
+                        continue
+                    yield pydantic.parse_obj_as(ExecutePromptEvent, json.loads(_text))  # type: ignore
+                return
+            _response.read()
+            if _response.status_code == 400:
+                raise BadRequestError(pydantic.parse_obj_as(typing.Any, _response.json()))  # type: ignore
+            if _response.status_code == 403:
+                raise ForbiddenError(pydantic.parse_obj_as(typing.Any, _response.json()))  # type: ignore
+            if _response.status_code == 404:
+                raise NotFoundError(pydantic.parse_obj_as(typing.Any, _response.json()))  # type: ignore
+            if _response.status_code == 500:
+                raise InternalServerError(pydantic.parse_obj_as(typing.Any, _response.json()))  # type: ignore
+            try:
+                _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)
+
     def execute_workflow_stream(
         self,
         *,
@@ -187,7 +352,7 @@ class Vellum:
         if _response.status_code == 400:
             raise BadRequestError(pydantic.parse_obj_as(typing.Any, _response.json()))  # type: ignore
         if _response.status_code == 403:
-            raise ForbiddenError(pydantic.parse_obj_as(GenerateErrorResponse, _response.json()))  # type: ignore
+            raise ForbiddenError(pydantic.parse_obj_as(typing.Any, _response.json()))  # type: ignore
         if _response.status_code == 404:
             raise NotFoundError(pydantic.parse_obj_as(typing.Any, _response.json()))  # type: ignore
         if _response.status_code == 500:
@@ -244,7 +409,7 @@ class Vellum:
             if _response.status_code == 400:
                 raise BadRequestError(pydantic.parse_obj_as(typing.Any, _response.json()))  # type: ignore
             if _response.status_code == 403:
-                raise ForbiddenError(pydantic.parse_obj_as(GenerateErrorResponse, _response.json()))  # type: ignore
+                raise ForbiddenError(pydantic.parse_obj_as(typing.Any, _response.json()))  # type: ignore
             if _response.status_code == 404:
                 raise NotFoundError(pydantic.parse_obj_as(typing.Any, _response.json()))  # type: ignore
             if _response.status_code == 500:
@@ -432,6 +597,167 @@ class AsyncVellum:
         self.sandboxes = AsyncSandboxesClient(client_wrapper=self._client_wrapper)
         self.test_suites = AsyncTestSuitesClient(client_wrapper=self._client_wrapper)
 
+    async def execute_prompt(
+        self,
+        *,
+        inputs: typing.List[PromptDeploymentInputRequest],
+        prompt_deployment_id: typing.Optional[str] = OMIT,
+        prompt_deployment_name: typing.Optional[str] = OMIT,
+        release_tag: typing.Optional[str] = OMIT,
+        external_id: typing.Optional[str] = OMIT,
+        expand_meta: typing.Optional[PromptDeploymentExpandMetaRequestRequest] = OMIT,
+        raw_overrides: typing.Optional[RawPromptExecutionOverridesRequest] = OMIT,
+        expand_raw: typing.Optional[typing.List[str]] = OMIT,
+        metadata: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+    ) -> ExecutePromptResponse:
+        """
+        Executes a deployed Prompt and returns the result.
+
+        Note: This endpoint temporarily does not support prompts with function calling, support is coming soon.
+        In the meantime, we recommend still using the `/generate` endpoint for prompts with function calling.
+
+        Parameters:
+            - inputs: typing.List[PromptDeploymentInputRequest].
+
+            - prompt_deployment_id: typing.Optional[str]. The ID of the Prompt Deployment. Must provide either this or prompt_deployment_name.
+
+            - prompt_deployment_name: typing.Optional[str]. The name of the Prompt Deployment. Must provide either this or prompt_deployment_id.
+
+            - release_tag: typing.Optional[str]. Optionally specify a release tag if you want to pin to a specific release of the Prompt Deployment
+
+            - external_id: typing.Optional[str].
+
+            - expand_meta: typing.Optional[PromptDeploymentExpandMetaRequestRequest]. The name of the Prompt Deployment. Must provide either this or prompt_deployment_id.
+
+            - raw_overrides: typing.Optional[RawPromptExecutionOverridesRequest].
+
+            - expand_raw: typing.Optional[typing.List[str]]. Returns the raw API response data sent from the model host. Combined with `raw_overrides`, it can be used to access new features from models.
+
+            - metadata: typing.Optional[typing.Dict[str, typing.Any]].
+        """
+        _request: typing.Dict[str, typing.Any] = {"inputs": inputs}
+        if prompt_deployment_id is not OMIT:
+            _request["prompt_deployment_id"] = prompt_deployment_id
+        if prompt_deployment_name is not OMIT:
+            _request["prompt_deployment_name"] = prompt_deployment_name
+        if release_tag is not OMIT:
+            _request["release_tag"] = release_tag
+        if external_id is not OMIT:
+            _request["external_id"] = external_id
+        if expand_meta is not OMIT:
+            _request["expand_meta"] = expand_meta
+        if raw_overrides is not OMIT:
+            _request["raw_overrides"] = raw_overrides
+        if expand_raw is not OMIT:
+            _request["expand_raw"] = expand_raw
+        if metadata is not OMIT:
+            _request["metadata"] = metadata
+        _response = await self._client_wrapper.httpx_client.request(
+            "POST",
+            urllib.parse.urljoin(f"{self._client_wrapper.get_environment().predict}/", "v1/execute-prompt"),
+            json=jsonable_encoder(_request),
+            headers=self._client_wrapper.get_headers(),
+            timeout=None,
+        )
+        if 200 <= _response.status_code < 300:
+            return pydantic.parse_obj_as(ExecutePromptResponse, _response.json())  # type: ignore
+        if _response.status_code == 400:
+            raise BadRequestError(pydantic.parse_obj_as(typing.Any, _response.json()))  # type: ignore
+        if _response.status_code == 403:
+            raise ForbiddenError(pydantic.parse_obj_as(typing.Any, _response.json()))  # type: ignore
+        if _response.status_code == 404:
+            raise NotFoundError(pydantic.parse_obj_as(typing.Any, _response.json()))  # type: ignore
+        if _response.status_code == 500:
+            raise InternalServerError(pydantic.parse_obj_as(typing.Any, _response.json()))  # type: ignore
+        try:
+            _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)
+
+    async def execute_prompt_stream(
+        self,
+        *,
+        inputs: typing.List[PromptDeploymentInputRequest],
+        prompt_deployment_id: typing.Optional[str] = OMIT,
+        prompt_deployment_name: typing.Optional[str] = OMIT,
+        release_tag: typing.Optional[str] = OMIT,
+        external_id: typing.Optional[str] = OMIT,
+        expand_meta: typing.Optional[PromptDeploymentExpandMetaRequestRequest] = OMIT,
+        raw_overrides: typing.Optional[RawPromptExecutionOverridesRequest] = OMIT,
+        expand_raw: typing.Optional[typing.List[str]] = OMIT,
+        metadata: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+    ) -> typing.AsyncIterator[ExecutePromptEvent]:
+        """
+        Executes a deployed Prompt and streams back the results.
+
+        Note: This endpoint temporarily does not support prompts with function calling, support is coming soon.
+        In the meantime, we recommend still using the `/generate-stream` endpoint for prompts with function calling
+
+        Parameters:
+            - inputs: typing.List[PromptDeploymentInputRequest].
+
+            - prompt_deployment_id: typing.Optional[str]. The ID of the Prompt Deployment. Must provide either this or prompt_deployment_name.
+
+            - prompt_deployment_name: typing.Optional[str]. The name of the Prompt Deployment. Must provide either this or prompt_deployment_id.
+
+            - release_tag: typing.Optional[str]. Optionally specify a release tag if you want to pin to a specific release of the Prompt Deployment
+
+            - external_id: typing.Optional[str].
+
+            - expand_meta: typing.Optional[PromptDeploymentExpandMetaRequestRequest]. The name of the Prompt Deployment. Must provide either this or prompt_deployment_id.
+
+            - raw_overrides: typing.Optional[RawPromptExecutionOverridesRequest].
+
+            - expand_raw: typing.Optional[typing.List[str]]. Returns the raw API response data sent from the model host. Combined with `raw_overrides`, it can be used to access new features from models.
+
+            - metadata: typing.Optional[typing.Dict[str, typing.Any]].
+        """
+        _request: typing.Dict[str, typing.Any] = {"inputs": inputs}
+        if prompt_deployment_id is not OMIT:
+            _request["prompt_deployment_id"] = prompt_deployment_id
+        if prompt_deployment_name is not OMIT:
+            _request["prompt_deployment_name"] = prompt_deployment_name
+        if release_tag is not OMIT:
+            _request["release_tag"] = release_tag
+        if external_id is not OMIT:
+            _request["external_id"] = external_id
+        if expand_meta is not OMIT:
+            _request["expand_meta"] = expand_meta
+        if raw_overrides is not OMIT:
+            _request["raw_overrides"] = raw_overrides
+        if expand_raw is not OMIT:
+            _request["expand_raw"] = expand_raw
+        if metadata is not OMIT:
+            _request["metadata"] = metadata
+        async with self._client_wrapper.httpx_client.stream(
+            "POST",
+            urllib.parse.urljoin(f"{self._client_wrapper.get_environment().predict}/", "v1/execute-prompt-stream"),
+            json=jsonable_encoder(_request),
+            headers=self._client_wrapper.get_headers(),
+            timeout=None,
+        ) as _response:
+            if 200 <= _response.status_code < 300:
+                async for _text in _response.aiter_lines():
+                    if len(_text) == 0:
+                        continue
+                    yield pydantic.parse_obj_as(ExecutePromptEvent, json.loads(_text))  # type: ignore
+                return
+            await _response.aread()
+            if _response.status_code == 400:
+                raise BadRequestError(pydantic.parse_obj_as(typing.Any, _response.json()))  # type: ignore
+            if _response.status_code == 403:
+                raise ForbiddenError(pydantic.parse_obj_as(typing.Any, _response.json()))  # type: ignore
+            if _response.status_code == 404:
+                raise NotFoundError(pydantic.parse_obj_as(typing.Any, _response.json()))  # type: ignore
+            if _response.status_code == 500:
+                raise InternalServerError(pydantic.parse_obj_as(typing.Any, _response.json()))  # type: ignore
+            try:
+                _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)
+
     async def execute_workflow_stream(
         self,
         *,
@@ -553,7 +879,7 @@ class AsyncVellum:
         if _response.status_code == 400:
             raise BadRequestError(pydantic.parse_obj_as(typing.Any, _response.json()))  # type: ignore
         if _response.status_code == 403:
-            raise ForbiddenError(pydantic.parse_obj_as(GenerateErrorResponse, _response.json()))  # type: ignore
+            raise ForbiddenError(pydantic.parse_obj_as(typing.Any, _response.json()))  # type: ignore
         if _response.status_code == 404:
             raise NotFoundError(pydantic.parse_obj_as(typing.Any, _response.json()))  # type: ignore
         if _response.status_code == 500:
@@ -610,7 +936,7 @@ class AsyncVellum:
             if _response.status_code == 400:
                 raise BadRequestError(pydantic.parse_obj_as(typing.Any, _response.json()))  # type: ignore
             if _response.status_code == 403:
-                raise ForbiddenError(pydantic.parse_obj_as(GenerateErrorResponse, _response.json()))  # type: ignore
+                raise ForbiddenError(pydantic.parse_obj_as(typing.Any, _response.json()))  # type: ignore
             if _response.status_code == 404:
                 raise NotFoundError(pydantic.parse_obj_as(typing.Any, _response.json()))  # type: ignore
             if _response.status_code == 500:

vellum/core/client_wrapper.py

@@ -16,7 +16,7 @@ class BaseClientWrapper:
         headers: typing.Dict[str, str] = {
             "X-Fern-Language": "Python",
             "X-Fern-SDK-Name": "vellum-ai",
-            "X-Fern-SDK-Version": "0.1.8",
+            "X-Fern-SDK-Version": "v0.1.10",
         }
         headers["X_API_KEY"] = self.api_key
         return headers

vellum/errors/forbidden_error.py

@@ -1,9 +1,10 @@
 # This file was auto-generated by Fern from our API Definition.
 
+import typing
+
 from ..core.api_error import ApiError
-from ..types.generate_error_response import GenerateErrorResponse
 
 
 class ForbiddenError(ApiError):
-    def __init__(self, body: GenerateErrorResponse):
+    def __init__(self, body: typing.Any):
         super().__init__(status_code=403, body=body)

vellum/resources/registered_prompts/client.py

@@ -63,6 +63,7 @@ class RegisteredPromptsClient:
                                                        * `HOSTED` - Hosted
                                                        * `MOSAICML` - MosaicML
                                                        * `OPENAI` - OpenAI
+                                                       * `FIREWORKS_AI` - Fireworks AI
                                                        * `HUGGINGFACE` - HuggingFace
                                                        * `MYSTIC` - Mystic
                                                        * `PYQ` - Pyq
@@ -146,6 +147,7 @@ class AsyncRegisteredPromptsClient:
                                                        * `HOSTED` - Hosted
                                                        * `MOSAICML` - MosaicML
                                                        * `OPENAI` - OpenAI
+                                                       * `FIREWORKS_AI` - Fireworks AI
                                                        * `HUGGINGFACE` - HuggingFace
                                                        * `MYSTIC` - Mystic
                                                        * `PYQ` - Pyq