A Guth Labs publication ↗
Back to News

Google adds page, cell and time-range selection to Gemini API file processing

Blank cream document layers beneath a green glass selection frame illuminated by a gold lamp
AI-generated editorial illustration

Google added selective file processing to the Gemini API on November 18, allowing developers to choose which parts of documents, spreadsheets, video and audio a model should examine, according to its release notes.

The controls cover page ranges, worksheet tabs, spreadsheet cell ranges and time intervals. Developers can also select content layers, including page images, extracted text, formulas and video transcripts. These choices work in user prompts and function responses, the results returned by tools an AI application calls.

Google says the change lets applications process and pay for only the content they need. Its document-processing guide describes inspecting specific pages, sheets or layers rather than an entire file, alongside support for Microsoft Office and OpenDocument formats.

This is not the API’s first support for office files: Google’s November 5 release notes already listed additional PowerPoint and OpenDocument formats. The new announcement focuses on controlling how files are processed.

For developers building document-review and media-analysis workflows, that distinction matters. Selecting the relevant material can reduce unnecessary input without requiring the model to reason over a whole workbook, presentation or recording.

Sources and citations

Revision history

Published .