api.md

Shared Types

from llama_cloud.types import (
    CloudAstraDBVectorStore,
    CloudAzStorageBlobDataSource,
    CloudAzureAISearchVectorStore,
    CloudBoxDataSource,
    CloudConfluenceDataSource,
    CloudGoogleDriveDataSource,
    CloudJiraDataSource,
    CloudJiraDataSourceV2,
    CloudMilvusVectorStore,
    CloudMongoDBAtlasVectorSearch,
    CloudNotionPageDataSource,
    CloudOneDriveDataSource,
    CloudPineconeVectorStore,
    CloudPostgresVectorStore,
    CloudQdrantVectorStore,
    CloudS3DataSource,
    CloudSharepointDataSource,
    CloudSlackDataSource,
    FailureHandlingConfig,
    PgVectorHnswSettings,
)

Files

Types:

from llama_cloud.types import (
    File,
    FileCreate,
    PresignedURL,
    FileCreateResponse,
    FileListResponse,
    FileQueryResponse,
)

Methods:

• client.files.create(**params) -> FileCreateResponse
• client.files.list(**params) -> SyncPaginatedCursor[FileListResponse]
• client.files.delete(file_id, **params) -> None
• client.files.get(file_id, **params) -> PresignedURL
• client.files.query(**params) -> FileQueryResponse

Parsing

Types:

from llama_cloud.types import (
    BBox,
    CodeItem,
    FailPageMode,
    FooterItem,
    HeaderItem,
    HeadingItem,
    ImageItem,
    LinkItem,
    ListItem,
    LlamaParseSupportedFileExtensions,
    ParsingJob,
    ParsingLanguages,
    ParsingMode,
    StatusEnum,
    TableItem,
    TextItem,
    ParsingCreateResponse,
    ParsingListResponse,
    ParsingGetResponse,
    ParsingUploadFileResponse,
)

Methods:

• client.parsing.create(**params) -> ParsingCreateResponse
• client.parsing.list(**params) -> SyncPaginatedCursor[ParsingListResponse]
• client.parsing.get(job_id, **params) -> ParsingGetResponse

Extraction

Methods:

• client.extraction.run(**params) -> ExtractJob

Jobs

Types:

from llama_cloud.types.extraction import (
    ExtractJob,
    WebhookConfiguration,
    JobListResponse,
    JobGetResultResponse,
)

Methods:

• client.extraction.jobs.create(**params) -> ExtractJob
• client.extraction.jobs.list(**params) -> JobListResponse
• client.extraction.jobs.file(**params) -> ExtractJob
• client.extraction.jobs.get(job_id) -> ExtractJob
• client.extraction.jobs.get_result(job_id, **params) -> JobGetResultResponse

Runs

Types:

from llama_cloud.types.extraction import ExtractConfig, ExtractRun

Methods:

• client.extraction.runs.list(**params) -> SyncPaginatedExtractRuns[ExtractRun]
• client.extraction.runs.delete(run_id, **params) -> object
• client.extraction.runs.get(run_id, **params) -> ExtractRun
• client.extraction.runs.get_by_job(job_id, **params) -> ExtractRun

ExtractionAgents

Types:

from llama_cloud.types.extraction import ExtractAgent, ExtractionAgentListResponse

Methods:

• client.extraction.extraction_agents.create(**params) -> ExtractAgent
• client.extraction.extraction_agents.update(extraction_agent_id, **params) -> ExtractAgent
• client.extraction.extraction_agents.list(**params) -> ExtractionAgentListResponse
• client.extraction.extraction_agents.delete(extraction_agent_id) -> object
• client.extraction.extraction_agents.get(extraction_agent_id) -> ExtractAgent

Schema

Types:

from llama_cloud.types.extraction.extraction_agents import (
    SchemaGenerateSchemaResponse,
    SchemaValidateSchemaResponse,
)

Methods:

• client.extraction.extraction_agents.schema.generate_schema(**params) -> SchemaGenerateSchemaResponse
• client.extraction.extraction_agents.schema.validate_schema(**params) -> SchemaValidateSchemaResponse

Classifier

Jobs

Types:

from llama_cloud.types.classifier import (
    ClassifierRule,
    ClassifyJob,
    ClassifyParsingConfiguration,
    JobGetResultsResponse,
)

Methods:

• client.classifier.jobs.create(**params) -> ClassifyJob
• client.classifier.jobs.list(**params) -> SyncPaginatedCursor[ClassifyJob]
• client.classifier.jobs.get(classify_job_id, **params) -> ClassifyJob
• client.classifier.jobs.get_results(classify_job_id, **params) -> JobGetResultsResponse

Projects

Types:

from llama_cloud.types import Project, ProjectListResponse

Methods:

• client.projects.list(**params) -> ProjectListResponse
• client.projects.get(project_id, **params) -> Project

DataSinks

Types:

from llama_cloud.types import DataSink, DataSinkListResponse

Methods:

• client.data_sinks.create(**params) -> DataSink
• client.data_sinks.update(data_sink_id, **params) -> DataSink
• client.data_sinks.list(**params) -> DataSinkListResponse
• client.data_sinks.delete(data_sink_id) -> None
• client.data_sinks.get(data_sink_id) -> DataSink

DataSources

Types:

from llama_cloud.types import DataSource, DataSourceReaderVersionMetadata, DataSourceListResponse

Methods:

• client.data_sources.create(**params) -> DataSource
• client.data_sources.update(data_source_id, **params) -> DataSource
• client.data_sources.list(**params) -> DataSourceListResponse
• client.data_sources.delete(data_source_id) -> None
• client.data_sources.get(data_source_id) -> DataSource

Pipelines

Types:

from llama_cloud.types import (
    AdvancedModeTransformConfig,
    AutoTransformConfig,
    AzureOpenAIEmbedding,
    AzureOpenAIEmbeddingConfig,
    BedrockEmbedding,
    BedrockEmbeddingConfig,
    CohereEmbedding,
    CohereEmbeddingConfig,
    DataSinkCreate,
    GeminiEmbedding,
    GeminiEmbeddingConfig,
    HuggingFaceInferenceAPIEmbedding,
    HuggingFaceInferenceAPIEmbeddingConfig,
    LlamaParseParameters,
    LlmParameters,
    ManagedIngestionStatusResponse,
    MessageRole,
    MetadataFilters,
    OpenAIEmbedding,
    OpenAIEmbeddingConfig,
    PageFigureNodeWithScore,
    PageScreenshotNodeWithScore,
    Pipeline,
    PipelineCreate,
    PipelineMetadataConfig,
    PipelineType,
    PresetRetrievalParams,
    RetrievalMode,
    SparseModelConfig,
    VertexAIEmbeddingConfig,
    VertexTextEmbedding,
    PipelineRetrieveResponse,
    PipelineListResponse,
)

Methods:

• client.pipelines.create(**params) -> Pipeline
• client.pipelines.retrieve(pipeline_id, **params) -> PipelineRetrieveResponse
• client.pipelines.update(pipeline_id, **params) -> Pipeline
• client.pipelines.list(**params) -> PipelineListResponse
• client.pipelines.delete(pipeline_id) -> None
• client.pipelines.get(pipeline_id) -> Pipeline
• client.pipelines.get_status(pipeline_id, **params) -> ManagedIngestionStatusResponse
• client.pipelines.upsert(**params) -> Pipeline

Sync

Methods:

• client.pipelines.sync.create(pipeline_id) -> Pipeline
• client.pipelines.sync.cancel(pipeline_id) -> Pipeline

DataSources

Types:

from llama_cloud.types.pipelines import (
    PipelineDataSource,
    DataSourceGetDataSourcesResponse,
    DataSourceUpdateDataSourcesResponse,
)

Methods:

• client.pipelines.data_sources.update(data_source_id, *, pipeline_id, **params) -> PipelineDataSource
• client.pipelines.data_sources.get_data_sources(pipeline_id) -> DataSourceGetDataSourcesResponse
• client.pipelines.data_sources.get_status(data_source_id, *, pipeline_id) -> ManagedIngestionStatusResponse
• client.pipelines.data_sources.sync(data_source_id, *, pipeline_id, **params) -> Pipeline
• client.pipelines.data_sources.update_data_sources(pipeline_id, **params) -> DataSourceUpdateDataSourcesResponse

Images

Types:

from llama_cloud.types.pipelines import (
    ImageListPageFiguresResponse,
    ImageListPageScreenshotsResponse,
)

Methods:

• client.pipelines.images.get_page_figure(figure_name, *, id, page_index, **params) -> object
• client.pipelines.images.get_page_screenshot(page_index, *, id, **params) -> object
• client.pipelines.images.list_page_figures(id, **params) -> ImageListPageFiguresResponse
• client.pipelines.images.list_page_screenshots(id, **params) -> ImageListPageScreenshotsResponse

Files

Types:

from llama_cloud.types.pipelines import (
    PipelineFile,
    FileCreateResponse,
    FileGetStatusCountsResponse,
)

Methods:

• client.pipelines.files.create(pipeline_id, **params) -> FileCreateResponse
• client.pipelines.files.update(file_id, *, pipeline_id, **params) -> PipelineFile
• client.pipelines.files.list(pipeline_id, **params) -> SyncPaginatedPipelineFiles[PipelineFile]
• client.pipelines.files.delete(file_id, *, pipeline_id) -> None
• client.pipelines.files.get_status(file_id, *, pipeline_id) -> ManagedIngestionStatusResponse
• client.pipelines.files.get_status_counts(pipeline_id, **params) -> FileGetStatusCountsResponse

Metadata

Types:

from llama_cloud.types.pipelines import MetadataCreateResponse

Methods:

• client.pipelines.metadata.create(pipeline_id, **params) -> MetadataCreateResponse
• client.pipelines.metadata.delete_all(pipeline_id) -> None

Documents

Types:

from llama_cloud.types.pipelines import (
    CloudDocument,
    CloudDocumentCreate,
    TextNode,
    DocumentCreateResponse,
    DocumentGetChunksResponse,
    DocumentUpsertResponse,
)

Methods:

• client.pipelines.documents.create(pipeline_id, **params) -> DocumentCreateResponse
• client.pipelines.documents.list(pipeline_id, **params) -> SyncPaginatedCloudDocuments[CloudDocument]
• client.pipelines.documents.delete(document_id, *, pipeline_id) -> None
• client.pipelines.documents.get(document_id, *, pipeline_id) -> CloudDocument
• client.pipelines.documents.get_chunks(document_id, *, pipeline_id) -> DocumentGetChunksResponse
• client.pipelines.documents.get_status(document_id, *, pipeline_id) -> ManagedIngestionStatusResponse
• client.pipelines.documents.sync(document_id, *, pipeline_id) -> object
• client.pipelines.documents.upsert(pipeline_id, **params) -> DocumentUpsertResponse

Retrievers

Types:

from llama_cloud.types import (
    CompositeRetrievalMode,
    CompositeRetrievalResult,
    ReRankConfig,
    Retriever,
    RetrieverCreate,
    RetrieverPipeline,
    RetrieverListResponse,
)

Methods:

• client.retrievers.create(**params) -> Retriever
• client.retrievers.update(retriever_id, **params) -> Retriever
• client.retrievers.list(**params) -> RetrieverListResponse
• client.retrievers.delete(retriever_id) -> None
• client.retrievers.get(retriever_id, **params) -> Retriever
• client.retrievers.search(**params) -> CompositeRetrievalResult
• client.retrievers.upsert(**params) -> Retriever

Retriever

Methods:

• client.retrievers.retriever.search(retriever_id, **params) -> CompositeRetrievalResult

Beta

AgentData

Types:

from llama_cloud.types.beta import (
    AgentData,
    AgentDataDeleteResponse,
    AgentDataAggregateResponse,
    AgentDataDeleteByQueryResponse,
)

Methods:

• client.beta.agent_data.update(item_id, **params) -> AgentData
• client.beta.agent_data.delete(item_id, **params) -> AgentDataDeleteResponse
• client.beta.agent_data.agent_data(**params) -> AgentData
• client.beta.agent_data.aggregate(**params) -> SyncPaginatedCursorPost[AgentDataAggregateResponse]
• client.beta.agent_data.delete_by_query(**params) -> AgentDataDeleteByQueryResponse
• client.beta.agent_data.get(item_id, **params) -> AgentData
• client.beta.agent_data.search(**params) -> SyncPaginatedCursorPost[AgentData]

ParseConfigurations

Types:

from llama_cloud.types.beta import (
    ParseConfiguration,
    ParseConfigurationCreate,
    ParseConfigurationQueryResponse,
)

Methods:

• client.beta.parse_configurations.create(**params) -> ParseConfiguration
• client.beta.parse_configurations.update(config_id, **params) -> ParseConfiguration
• client.beta.parse_configurations.list(**params) -> SyncPaginatedCursor[ParseConfiguration]
• client.beta.parse_configurations.delete(config_id, **params) -> None
• client.beta.parse_configurations.get(config_id, **params) -> ParseConfiguration

Sheets

Types:

from llama_cloud.types.beta import SheetsJob, SheetsParsingConfig

Methods:

• client.beta.sheets.create(**params) -> SheetsJob
• client.beta.sheets.list(**params) -> SyncPaginatedCursor[SheetsJob]
• client.beta.sheets.delete_job(spreadsheet_job_id, **params) -> object
• client.beta.sheets.get(spreadsheet_job_id, **params) -> SheetsJob
• client.beta.sheets.get_result_table(region_type, *, spreadsheet_job_id, region_id, **params) -> PresignedURL

Directories

Types:

from llama_cloud.types.beta import (
    DirectoryCreateResponse,
    DirectoryUpdateResponse,
    DirectoryListResponse,
    DirectoryGetResponse,
)

Methods:

• client.beta.directories.create(**params) -> DirectoryCreateResponse
• client.beta.directories.update(directory_id, **params) -> DirectoryUpdateResponse
• client.beta.directories.list(**params) -> SyncPaginatedCursor[DirectoryListResponse]
• client.beta.directories.delete(directory_id, **params) -> None
• client.beta.directories.get(directory_id, **params) -> DirectoryGetResponse

Files

Types:

from llama_cloud.types.beta.directories import (
    FileUpdateResponse,
    FileListResponse,
    FileAddResponse,
    FileGetResponse,
    FileUploadResponse,
)

Methods:

• client.beta.directories.files.update(directory_file_id, *, path_directory_id, **params) -> FileUpdateResponse
• client.beta.directories.files.list(directory_id, **params) -> SyncPaginatedCursor[FileListResponse]
• client.beta.directories.files.delete(directory_file_id, *, directory_id, **params) -> None
• client.beta.directories.files.add(directory_id, **params) -> FileAddResponse
• client.beta.directories.files.get(directory_file_id, *, directory_id, **params) -> FileGetResponse
• client.beta.directories.files.upload(directory_id, **params) -> FileUploadResponse

Batch

Types:

from llama_cloud.types.beta import (
    BatchCreateResponse,
    BatchListResponse,
    BatchCancelResponse,
    BatchGetStatusResponse,
)

Methods:

• client.beta.batch.create(**params) -> BatchCreateResponse
• client.beta.batch.list(**params) -> SyncPaginatedBatchItems[BatchListResponse]
• client.beta.batch.cancel(job_id, **params) -> BatchCancelResponse
• client.beta.batch.get_status(job_id, **params) -> BatchGetStatusResponse

JobItems

Types:

from llama_cloud.types.beta.batch import JobItemListResponse, JobItemGetProcessingResultsResponse

Methods:

• client.beta.batch.job_items.list(job_id, **params) -> SyncPaginatedBatchItems[JobItemListResponse]
• client.beta.batch.job_items.get_processing_results(item_id, **params) -> JobItemGetProcessingResultsResponse

Split

Types:

from llama_cloud.types.beta import (
    SplitCategory,
    SplitDocumentInput,
    SplitResultResponse,
    SplitSegmentResponse,
    SplitCreateResponse,
    SplitListResponse,
    SplitGetResponse,
)

Methods:

• client.beta.split.create(**params) -> SplitCreateResponse
• client.beta.split.list(**params) -> SyncPaginatedCursor[SplitListResponse]
• client.beta.split.get(split_job_id, **params) -> SplitGetResponse