LlamaCloud, documented through the LlamaParse developer docs, offers hosted APIs for getting data out of documents. Parse converts PDFs, scans and Office files into markdown or JSON, while Extract returns JSON that follows your own schema.
Classify, Split and Index handle sorting, segmenting and retrieval. Everything works with one API key, and SDKs are available for Python, TypeScript, Go and Java, with integrations such as n8n and MCP.