Gemini API: Expanded Office and OpenDocument File Support
Gemini CLIThe Gemini API now accepts PPT, ODP and ODS files as input and adds page-image processing for DOC files. DOC, PPT and ODP files are rendered as page images for the model, while ODS files give the model the data and formulas from the first sheet. Handling for DOCX, DOTX, ODT, PPTX, XLSX and XLS is unchanged.
Key Takeaways
- PPT, ODP and ODS files are now accepted as Gemini API input, widening the set of documents sent without conversion.
- DOC files gain page-image processing, so the model can read their visual layout.
- DOC, PPT and ODP files are rendered as page images, which preserves slides, charts and formatting for the model.
- ODS spreadsheets expose only the first sheet, with its data and formulas, so multi-sheet files need preparation.
- Existing formats are untouched: DOCX, DOTX, ODT, PPTX, XLSX and XLS behave as before.
What Is New
The Gemini API expanded the document formats it accepts as input. Developers can now send PPT, ODP and ODS files directly, and DOC files gain page-image processing. This reduces the need to convert legacy Office files and OpenDocument files to PDF before sending them to a model.
How Each Format Is Handled
DOC, PPT and ODP files are rendered as page images, so the model sees each page or slide visually, including layout, charts and diagrams. ODS spreadsheets are handled differently: the model receives the data and formulas from the first sheet rather than page images. Developers whose ODS files keep important content on later sheets should split or reorder the file before sending it.
What Stays the Same
Handling for DOCX, DOTX, ODT, PPTX, XLSX and XLS files is unchanged, so existing integrations that use those formats need no modification.