AI agents for LlamaParse

Get a parsed document

Returns a parse job's job.status and, once COMPLETED, its content.

List parse jobs

Lists the project's parse jobs, newest first, with each one's id, name, status, tier and created_at.

Get an extraction

Returns an extraction job's status and, once COMPLETED, extract_result (the values in your schema's shape) and extract_metadata (citations and confidence when asked for).

Get a classification

Returns a classify job's status and, once COMPLETED, result.type (the matched type, or null when no rule fits), result.confidence and result.reasoning.

Parse a document

Starts reading a document from its public web address (PDF, Word, PowerPoint, Excel, HTML, images and more) into Markdown and text.

Parse a file

Uploads a document file (a person's attachment or another tool's answer, passed as its file reference fil_…) and starts reading it into Markdown and text, like parse_document does for a web address.

Extract structured data

Pulls the fields you describe out of a document and returns them as JSON that follows your schema: an invoice's number, dates, totals and line items, a contract's parties and terms.

Classify a document

Decides which of your document types a parsed document is.