* Delete all matching agent data, returns the total number of deleted items

Run stateless extraction and wait for results (synchronous version). Args: data_schema: The schema defining what data to extract

Context manager to add helpful information for errors due to nested event loops.

Determine MIME type for a file based on filename or path.

Block until the requested ingestion work is finished. - If `file_ids` is given → wait for those files. - If `doc_ids` is giv

Format markdown text for Jupyter notebook display by escaping dollar signs.

Parse the file and return a JobResult object instead of Document objects. This method is similar to aload_data but returns JobResult

Calculate the overall confidence score for the extracted data.

Load data from the input path. File(s) which were partitioned before parsing will be loaded as a single re-assembled Document.

Block until the requested ingestion work is finished. - If `file_ids` is given → wait for those files. - If `doc_ids` is giv

Run a command, streaming output to the console, and raise on failure.

Return a Retriever for this managed index.

Convert raw API response to typed agent data. Args: raw_data: Raw agent data from the API validator: Pydanti

Make an retrying API request to the LlamaParse API. Args: client: The httpx.AsyncClient to use for the request. url: The URL to r

Run coroutine in a separate thread to avoid event loop issues

Create a new LlamaCloud managed index.

* Loads data from a file and returns an array of ParseResult objects. * To be used with resultType "json". * * @param filePathOrContent - Th

Load data from the input path.

Run coroutine in a separate thread to avoid event loop issues

Download a region result (either region data or cell metadata). Args: job_id: ID of the job region_id: ID of the regi

Build a LlamaCloud managed index from a sequence of documents.

Create an ExtractedData instance from an extraction result.

* Loads data from a file and returns an array of Document objects. * To be used with resultType "text" or "markdown". * * @param fileContent

Yield partitioned target_pages segments.

Classify a list of files by their IDs. Note that even if a job fails, some of the files may have been classified successfully.

Run stateless extraction and wait for results. Args: data_schema: The schema defining what data to extract config: Th

High-level method to parse a spreadsheet file. This method handles the entire workflow: 1. Upload the file 2. Create a parsin

Check if an SDK update is available. Args: client: HTTPX client to use. quiet: If False, update availability will also be printed

Create a new LlamaCloud managed index.

Delete an extraction agent by ID. Args: agent_id (str): ID of the extraction agent to delete

Yield all values in target_pages.

Parse the file and return a JobResult object instead of Document objects. This method is similar to load_data but returns JobResult

Create a new ExtractedData instance with sensible defaults. Args: extracted_data: The extracted data payload

Get extraction agents by name or extraction agent ID. Args: name (Optional[str]): Filter by name extraction_agent_id

List all available extraction agents.

List extraction runs for the extraction agent. Returns: PaginatedExtractRunsResponse: Paginated list of extraction runs

* Transform API response to typed data

Upload content from various input types or fetch an already-uploaded file. Args: file_input: The content to upload. Can

Upload a file from a URL to the index.

Get or create the async httpx client

Get default query parameters for API requests

Get common headers for API requests

Get job result with error handling based on ignore_errors setting.

Create one parse job and wait for the result.

Classify one or more files from various input types. Args: rules: The rules to use for classification. files

Get image data by name using the job ID. Args: image_name: The name of the image to fetch Returns:

Get the image documents from the job. Args: include_screenshot_images (bool): Whether to include screens

Load data from the input path.

Save an image to a file. Args: image_name: The name of the image to fetch output_dir: The directory to save

Create a new extraction agent. Args: name (str): The name of the extraction agent data_schema (SchemaInput): The data

Get the markdown documents from the job. Args: split_by_page: Whether to split the pages into separate documents

Get the documents from the job. Args: split_by_page: Whether to split the pages into separate documents

Check if we're running in a Jupyter environment.

Queue extraction jobs using stateless extraction (no agent required). Args: data_schema: The schema defining what data to extract

Persist the extraction agent's schema and config to the database. Returns: ExtractionAgent: The updated extraction agent

* Splits text into Document objects using the page separator. * * @param text - The text to be split. * @returns An array of Document objects

* Creates a job for the LlamaParse API. * * @param data - The file data as a Uint8Array. * @param filename - Optional filename for the file.

Load data from the input path.

Load data from the input path.

Poll `get_status_fn` until every id in `resource_ids` is finished. Args: resource_ids: Iterable of resource ids to watch

Calculate the next backoff interval based on the backoff pattern. Args: current_interval: The current interval in seconds

Convert FileInput to FileData or text string for stateless extraction.

Upload a file from various input types using FileClient.