api.md

Shared Types

from openai.types import (
    ErrorObject,
    FunctionDefinition,
    FunctionParameters,
    ResponseFormatJSONObject,
    ResponseFormatJSONSchema,
    ResponseFormatText,
)

Completions

Types:

from openai.types import Completion, CompletionChoice, CompletionUsage

Methods:

• client.completions.create(**params) -> Completion

Chat

Types:

from openai.types import ChatModel

Completions

Types:

from openai.types.chat import (
    ChatCompletion,
    ChatCompletionAssistantMessageParam,
    ChatCompletionChunk,
    ChatCompletionContentPart,
    ChatCompletionContentPartImage,
    ChatCompletionContentPartRefusal,
    ChatCompletionContentPartText,
    ChatCompletionFunctionCallOption,
    ChatCompletionFunctionMessageParam,
    ChatCompletionMessage,
    ChatCompletionMessageParam,
    ChatCompletionMessageToolCall,
    ChatCompletionNamedToolChoice,
    ChatCompletionRole,
    ChatCompletionStreamOptions,
    ChatCompletionSystemMessageParam,
    ChatCompletionTokenLogprob,
    ChatCompletionTool,
    ChatCompletionToolChoiceOption,
    ChatCompletionToolMessageParam,
    ChatCompletionUserMessageParam,
)

Methods:

• client.chat.completions.create(**params) -> ChatCompletion

Embeddings

Types:

from openai.types import CreateEmbeddingResponse, Embedding

Methods:

• client.embeddings.create(**params) -> CreateEmbeddingResponse

Files

Types:

from openai.types import FileContent, FileDeleted, FileObject, FilePurpose

Methods:

• client.files.create(**params) -> FileObject
• client.files.retrieve(file_id) -> FileObject
• client.files.list(**params) -> SyncPage[FileObject]
• client.files.delete(file_id) -> FileDeleted
• client.files.content(file_id) -> HttpxBinaryResponseContent
• client.files.retrieve_content(file_id) -> str
• client.files.wait_for_processing(*args) -> FileObject

Images

Types:

from openai.types import Image, ImageModel, ImagesResponse

Methods:

• client.images.create_variation(**params) -> ImagesResponse
• client.images.edit(**params) -> ImagesResponse
• client.images.generate(**params) -> ImagesResponse

Audio

Types:

from openai.types import AudioModel

Transcriptions

Types:

from openai.types.audio import Transcription

Methods:

• client.audio.transcriptions.create(**params) -> Transcription

Translations

Types:

from openai.types.audio import Translation

Methods:

• client.audio.translations.create(**params) -> Translation

Speech

Types:

from openai.types.audio import SpeechModel

Methods:

• client.audio.speech.create(**params) -> HttpxBinaryResponseContent

Moderations

Types:

from openai.types import Moderation, ModerationModel, ModerationCreateResponse

Methods:

• client.moderations.create(**params) -> ModerationCreateResponse

Models

Types:

from openai.types import Model, ModelDeleted

Methods:

• client.models.retrieve(model) -> Model
• client.models.list() -> SyncPage[Model]
• client.models.delete(model) -> ModelDeleted

FineTuning

Jobs

Types:

from openai.types.fine_tuning import (
    FineTuningJob,
    FineTuningJobEvent,
    FineTuningJobIntegration,
    FineTuningJobWandbIntegration,
    FineTuningJobWandbIntegrationObject,
)

Methods:

• client.fine_tuning.jobs.create(**params) -> FineTuningJob
• client.fine_tuning.jobs.retrieve(fine_tuning_job_id) -> FineTuningJob
• client.fine_tuning.jobs.list(**params) -> SyncCursorPage[FineTuningJob]
• client.fine_tuning.jobs.cancel(fine_tuning_job_id) -> FineTuningJob
• client.fine_tuning.jobs.list_events(fine_tuning_job_id, **params) -> SyncCursorPage[FineTuningJobEvent]

Checkpoints

Types:

from openai.types.fine_tuning.jobs import FineTuningJobCheckpoint

Methods:

• client.fine_tuning.jobs.checkpoints.list(fine_tuning_job_id, **params) -> SyncCursorPage[FineTuningJobCheckpoint]

Beta

VectorStores

Types:

from openai.types.beta import VectorStore, VectorStoreDeleted

Methods:

• client.beta.vector_stores.create(**params) -> VectorStore
• client.beta.vector_stores.retrieve(vector_store_id) -> VectorStore
• client.beta.vector_stores.update(vector_store_id, **params) -> VectorStore
• client.beta.vector_stores.list(**params) -> SyncCursorPage[VectorStore]
• client.beta.vector_stores.delete(vector_store_id) -> VectorStoreDeleted

Files

Types:

from openai.types.beta.vector_stores import VectorStoreFile, VectorStoreFileDeleted

Methods:

• client.beta.vector_stores.files.create(vector_store_id, **params) -> VectorStoreFile
• client.beta.vector_stores.files.retrieve(file_id, *, vector_store_id) -> VectorStoreFile
• client.beta.vector_stores.files.list(vector_store_id, **params) -> SyncCursorPage[VectorStoreFile]
• client.beta.vector_stores.files.delete(file_id, *, vector_store_id) -> VectorStoreFileDeleted
• client.beta.vector_stores.files.create_and_poll(*args) -> VectorStoreFile
• client.beta.vector_stores.files.poll(*args) -> VectorStoreFile
• client.beta.vector_stores.files.upload(*args) -> VectorStoreFile
• client.beta.vector_stores.files.upload_and_poll(*args) -> VectorStoreFile

FileBatches

Types:

from openai.types.beta.vector_stores import VectorStoreFileBatch

Methods:

• client.beta.vector_stores.file_batches.create(vector_store_id, **params) -> VectorStoreFileBatch
• client.beta.vector_stores.file_batches.retrieve(batch_id, *, vector_store_id) -> VectorStoreFileBatch
• client.beta.vector_stores.file_batches.cancel(batch_id, *, vector_store_id) -> VectorStoreFileBatch
• client.beta.vector_stores.file_batches.list_files(batch_id, *, vector_store_id, **params) -> SyncCursorPage[VectorStoreFile]
• client.beta.vector_stores.file_batches.create_and_poll(*args) -> VectorStoreFileBatch
• client.beta.vector_stores.file_batches.poll(*args) -> VectorStoreFileBatch
• client.beta.vector_stores.file_batches.upload_and_poll(*args) -> VectorStoreFileBatch

Assistants

Types:

from openai.types.beta import (
    Assistant,
    AssistantDeleted,
    AssistantStreamEvent,
    AssistantTool,
    CodeInterpreterTool,
    FileSearchTool,
    FunctionTool,
    MessageStreamEvent,
    RunStepStreamEvent,
    RunStreamEvent,
    ThreadStreamEvent,
)

Methods:

• client.beta.assistants.create(**params) -> Assistant
• client.beta.assistants.retrieve(assistant_id) -> Assistant
• client.beta.assistants.update(assistant_id, **params) -> Assistant
• client.beta.assistants.list(**params) -> SyncCursorPage[Assistant]
• client.beta.assistants.delete(assistant_id) -> AssistantDeleted

Threads

Types:

from openai.types.beta import (
    AssistantResponseFormatOption,
    AssistantToolChoice,
    AssistantToolChoiceFunction,
    AssistantToolChoiceOption,
    Thread,
    ThreadDeleted,
)

Methods:

• client.beta.threads.create(**params) -> Thread
• client.beta.threads.retrieve(thread_id) -> Thread
• client.beta.threads.update(thread_id, **params) -> Thread
• client.beta.threads.delete(thread_id) -> ThreadDeleted
• client.beta.threads.create_and_run(**params) -> Run
• client.beta.threads.create_and_run_poll(*args) -> Run
• client.beta.threads.create_and_run_stream(*args) -> AssistantStreamManager[AssistantEventHandler] | AssistantStreamManager[AssistantEventHandlerT]

Runs

Types:

from openai.types.beta.threads import RequiredActionFunctionToolCall, Run, RunStatus

Methods:

• client.beta.threads.runs.create(thread_id, **params) -> Run
• client.beta.threads.runs.retrieve(run_id, *, thread_id) -> Run
• client.beta.threads.runs.update(run_id, *, thread_id, **params) -> Run
• client.beta.threads.runs.list(thread_id, **params) -> SyncCursorPage[Run]
• client.beta.threads.runs.cancel(run_id, *, thread_id) -> Run
• client.beta.threads.runs.submit_tool_outputs(run_id, *, thread_id, **params) -> Run
• client.beta.threads.runs.create_and_poll(*args) -> Run
• client.beta.threads.runs.create_and_stream(*args) -> AssistantStreamManager[AssistantEventHandler] | AssistantStreamManager[AssistantEventHandlerT]
• client.beta.threads.runs.poll(*args) -> Run
• client.beta.threads.runs.stream(*args) -> AssistantStreamManager[AssistantEventHandler] | AssistantStreamManager[AssistantEventHandlerT]
• client.beta.threads.runs.submit_tool_outputs_and_poll(*args) -> Run
• client.beta.threads.runs.submit_tool_outputs_stream(*args) -> AssistantStreamManager[AssistantEventHandler] | AssistantStreamManager[AssistantEventHandlerT]

Steps

Types:

from openai.types.beta.threads.runs import (
    CodeInterpreterLogs,
    CodeInterpreterOutputImage,
    CodeInterpreterToolCall,
    CodeInterpreterToolCallDelta,
    FileSearchToolCall,
    FileSearchToolCallDelta,
    FunctionToolCall,
    FunctionToolCallDelta,
    MessageCreationStepDetails,
    RunStep,
    RunStepDelta,
    RunStepDeltaEvent,
    RunStepDeltaMessageDelta,
    RunStepInclude,
    ToolCall,
    ToolCallDelta,
    ToolCallDeltaObject,
    ToolCallsStepDetails,
)

Methods:

• client.beta.threads.runs.steps.retrieve(step_id, *, thread_id, run_id, **params) -> RunStep
• client.beta.threads.runs.steps.list(run_id, *, thread_id, **params) -> SyncCursorPage[RunStep]

Messages

Types:

from openai.types.beta.threads import (
    Annotation,
    AnnotationDelta,
    FileCitationAnnotation,
    FileCitationDeltaAnnotation,
    FilePathAnnotation,
    FilePathDeltaAnnotation,
    ImageFile,
    ImageFileContentBlock,
    ImageFileDelta,
    ImageFileDeltaBlock,
    ImageURL,
    ImageURLContentBlock,
    ImageURLDelta,
    ImageURLDeltaBlock,
    Message,
    MessageContent,
    MessageContentDelta,
    MessageContentPartParam,
    MessageDeleted,
    MessageDelta,
    MessageDeltaEvent,
    RefusalContentBlock,
    RefusalDeltaBlock,
    Text,
    TextContentBlock,
    TextContentBlockParam,
    TextDelta,
    TextDeltaBlock,
)

Methods:

• client.beta.threads.messages.create(thread_id, **params) -> Message
• client.beta.threads.messages.retrieve(message_id, *, thread_id) -> Message
• client.beta.threads.messages.update(message_id, *, thread_id, **params) -> Message
• client.beta.threads.messages.list(thread_id, **params) -> SyncCursorPage[Message]
• client.beta.threads.messages.delete(message_id, *, thread_id) -> MessageDeleted

Batches

Types:

from openai.types import Batch, BatchError, BatchRequestCounts

Methods:

• client.batches.create(**params) -> Batch
• client.batches.retrieve(batch_id) -> Batch
• client.batches.list(**params) -> SyncCursorPage[Batch]
• client.batches.cancel(batch_id) -> Batch

Uploads

Types:

from openai.types import Upload

Methods:

• client.uploads.create(**params) -> Upload
• client.uploads.cancel(upload_id) -> Upload
• client.uploads.complete(upload_id, **params) -> Upload

Parts

Types:

from openai.types.uploads import UploadPart

Methods:

• client.uploads.parts.create(upload_id, **params) -> UploadPart