PDF Vector is a hosted document parsing service, sold as an API (an interface that another program calls over the internet). It turns PDF, Word, Excel, PowerPoint, image and more than 20 other file types into Markdown, a plain-text format that keeps headings, lists and tables, a form that language models handle well.
Its 19 APIs follow one pattern. Parse returns the text of a document; Ask answers a question about it; Extract fills a JSON Schema (a description of the fields and types wanted) with data from it. Specialized versions handle invoices, identity documents and bank statements. Academic APIs search and fetch papers from PubMed, arXiv, OpenAlex and Semantic Scholar. The service says it stores no documents or results, offers dedicated instances in chosen regions, and connects to Zapier, Make, n8n and AI apps through an MCP server.
PDF Vector is a paid hosted service with a free tier. Every plan includes all APIs; plans differ by the number of monthly credits, which API calls consume.