Gemini Can Now Generate, Preview and Edit Files Inside the Chat
Google has opened file generation, preview and editing in its Gemini app to all Gemini users. The feature covers PDFs, Google Docs, Google Sheets, slide decks, Markdown files, TXT files and code, and it runs directly inside the Gemini interface.
Platform support is uneven. PDF preview and editing work on both web and mobile, while editing of the other formats is limited to the web version. For someone reviewing a contract on a phone, that is a practical difference. A spreadsheet or slide deck still requires a desktop browser.
Editing is scoped to selections rather than whole files. Users can expand a document to full-screen mode, then highlight text to ask questions or request changes. Think of it as a sticky note placed on one paragraph, not a request to rewrite the whole page. The Verge reports that one example is turning a highlighted table into a graph. That pattern, a natural-language instruction anchored to a region of an existing file, differs from regenerating an entire file from a prompt.
The release builds on several earlier file capabilities. According to Google's Gemini blog, Gemini can create Microsoft Word documents from a prompt and Microsoft Excel files directly in chat. Canvas, Gemini's workspace for drafting, lets users create, edit and share documents and code with Gemini in real time. In October 2025, the Gemini app added LaTeX-based PDF generation (LaTeX is a markup system widely used for technical and academic documents), which lets users generate a PDF, preview it in Gemini and download it for sharing or printing. Deep Research accepts uploaded PDFs and images, and its reports can be imported into Canvas for further refinement. In Google Drive, the Gemini side panel can summarize one or several documents and interact with PDFs. Google Docs added a feature to listen to documents using Gemini in August 2025. Google Vids can also turn Google Docs, PDFs and Word files into video summaries, using AI to generate scripts and narration with custom visuals.
The broader context here is consolidation. For much of the chatbot era, generated output arrived as text that users copied into whichever tool came next. Gemini's file handling brings the artifact itself into the conversation, so generation, review and revision happen in one place.
In my view, the questions that will matter most to enterprise administrators are less about whether the output reads well and more about how edited files are versioned, permissioned and returned to shared storage. Those details tend to decide whether a feature survives a compliance review.
The practical upside is a shorter path from question to finished file. A user can upload a report, ask for the figures to be charted, adjust one table and export a PDF without leaving the thread. For teams that already work in Google Workspace, that reduction in steps is the part most likely to change daily habits.
