Office documents, Excel files and rich content

Word, PowerPoint, Excel, OpenDocument, and EPUB files are turned into text an assistant can read. Layout is simplified: what matters is the wording, tables, and (when possible) figures.

DOCX is read directly. DOC and ODT are converted first, then read the same way. Headings, lists, and body text come through. Embedded figures follow the same image rules as in How Raison reads images inside your documents.

PPTX and ODP are converted, then read like a PDF, including text on slides. If that conversion cannot finish, Raison falls back to a simpler text extract. Image-heavy decks may need a second pass (the same option offered on scans).

XLSX is extracted as text and tables. Named sheets and cell values are what the assistant will see, not formulas, charts, or macros. Very wide sheets work better if you keep headers clear and drop unused columns before upload.

EPUB books are accepted. HTML and Markdown upload as text files. SCORM and H5P are packages, not Office files: see Import a SCORM package.

Need help? Write to [email protected].