Gemini adds in-chat editor for documents and spreadsheets

Streamlined document workflows

Google has introduced a new way for users to interact with files generated by the Gemini app. Users can now preview and edit documents, PDFs, slide decks, and spreadsheets without ever leaving the chat window. This update arrives six months after the company first enabled the ability to produce downloadable files.

Previously, the workflow required users to download a file, open it in a separate program like Word or Excel, and then return to the chat to request further changes. The new in-chat editor removes these steps by allowing for immediate adjustments. This is particularly useful for drafting work documents like campaign briefs or media plans.

The update is available to all users across every language supported by the Gemini app. While the feature is live for everyone, Google has not specified which specific AI model handles the editing requests. It is also unclear if these editing actions count toward the usage limits of paid subscription tiers.

Three ways to interact with files

The new interface provides three distinct methods for refining content once a file appears in the conversation. Users can expand a file to a full-screen view to type edits directly, much like a standard word processor. This allows for traditional manual control over the final output.

Alternatively, users can highlight a specific passage to turn a selection into a prompt. This allows for targeted questions about a single section of a document rather than the entire file. The third method hands the rewriting back to the AI, allowing Gemini to refine the content based on a specific instruction.

Google demonstrated these capabilities using a study guide scenario. A user can upload handwritten lecture notes, and Gemini can generate a PDF study guide from them. From there, the user can manually edit the text or highlight specific sections to have the AI adjust visuals or specific paragraphs.

Platform and format limitations

There is a notable split in functionality between mobile and web platforms. PDF preview and editing are supported on both mobile and web devices. However, other formats like Google Docs, Google Sheets, slide decks, Markdown files, and code are currently limited to the web version of the app.

This means that while phone users can still receive spreadsheets or slide decks from Gemini, they cannot perform in-chat edits on those specific formats on a mobile device. Users seeking to modify these files would need to switch to a web browser to access the full suite of editing tools.

The update also includes new export options for professional use. For example, slide decks generated in the chat can now be downloaded as Microsoft PowerPoint files. This allows teams to move content out of the Google ecosystem more easily than was possible in previous versions of the tool.

Impact on business and marketing

This feature reaches a massive audience, as Alphabet recently reported that the Gemini app has surpassed one billion monthly active users. For marketing agencies and corporate teams, this change affects the everyday output of their work, such as trackers, tables, and client presentations. It positions Gemini as a more robust productivity tool for finishing tasks.

The rollout is intended for all Google Workspace customers, including personal accounts and business subscribers. Administrators for Workspace domains can choose between Rapid and Scheduled release tracks, though there is no specific toggle to disable file generation while keeping the rest of the app active. Organizations must choose to enable or disable the Gemini app as a whole.

By integrating these tools, Google is competing directly with other AI workspaces like ChatGPT's canvas or Anthropic's spreadsheet capabilities. The goal is to keep users inside the Gemini interface for longer periods. Currently, this time is spent on productivity rather than being sold to advertisers, as Google maintains that the app remains ad-free.

Technical details and accuracy

The ability to process handwritten notes highlights Gemini's multimodal capabilities. This involves reading cursive, abbreviations, and diagrams from photographs, which is a more complex task than standard optical character recognition. However, because these inputs are harder to read, there is a risk of errors flowing into the final document.

Google has not stated whether the AI marks passages it is unsure about or provides a way to trace a specific sentence back to the original notes. Because a polished PDF can make errors harder to spot than a plain text reply, users should remain vigilant for hallucinations. Hallucinations occur when an AI produces confident but incorrect statements not found in the source material.

The help documentation notes that users can upload up to 10 files in a single prompt, with a size limit of 100 MB per non-video file. While the company has not listed specific page limits or caps on the number of files that can be generated, rolling limits govern how much data can be analyzed over a period of time.