Google Gemini Now Edits PDFs, Slides and Spreadsheets in Chat: What Works and What to Check

29 October 2026 | Official announcement: 29 October 2026 | Status: Google says available to all users in languages and locations where the Gemini app is available. A familiar AI workflow is to ask for a report, copy the response, open another application, fix the formatting, and repeat. Google's latest Gemini app update aims to remove some of those handoffs: users can now generate, preview and directly edit multiple file types inside the chat experience. The important detail is not merely that Gemini can write text. It is that Google now explicitly supports PDF preview and editing on web and mobile, with direct editing for other file formats on the web. This is the distinction in Google's dated 29 October product announcement.

What changed on 29 October?

Google says a user can generate a file, expand it into a full-screen view, type edits directly, select text to request a targeted change, or ask Gemini to refine the content. Supported examples include PDFs, Google Docs, Google Sheets, slide decks, Markdown, TXT and code. The announcement describes file generation, preview and editing together, rather than requiring every iteration to be a new chat response. Google also illustrates a PDF study guide made from uploaded handwritten notes, followed by manual and AI-assisted revisions. The source is the official Gemini app release post; Google does not publish a complete per-format fidelity or feature-parity matrix in that post.

What works on a phone, and what still needs the web?

PDFs: Google's announcement explicitly supports previewing and editing PDFs on both the Gemini web experience and mobile. Other file types: Google Docs, Sheets, slide decks, Markdown, TXT and code are listed as supported on the web only. That is a real workflow limitation: do not promise a mobile user that a spreadsheet or slide deck can be edited in the same way as a PDF. Google states the new capabilities are available to all users in languages where the Gemini app is available; this is not a promise that every organization, device configuration, policy or regional account has identical controls. Confirm the actual experience with the account that will use it. Source: Google's availability and platform statement.

Is this the first time Gemini could edit a document?

No. In March 2025, Google introduced Canvas as a collaborative space for editing documents and code, including highlighting a passage and requesting a change. Its 18 March 2025 Canvas announcement describes document editing and export to Google Docs. Google's current Canvas help documentation also describes direct edits, selection-based changes and version navigation. The 29 October story should therefore not be marketed as the invention of AI document editing. Its narrower, newsworthy claim is the newly announced direct file creation, preview and editing experience across multiple file formats, especially PDFs, with explicit web-versus-mobile support.

Three useful workflows worth testing

HR policy review: Start with a fictional leave-policy PDF, ask Gemini for a clearer employee-facing explanation, then select a paragraph to correct ambiguous wording. Check that dates, eligibility rules, exceptions and legal terms remain intact. Business proposal: Create a slide outline and refine its headings in the web experience, then verify the exported presentation in the application used for final delivery. Data summary: Ask for a small spreadsheet-style analysis from synthetic data and manually check every formula, total, chart label and locale-specific number format. These are suggested pilots, not vendor-confirmed accuracy guarantees. The Google announcement establishes the editing interface; it does not establish that complex business documents retain perfect formatting or formulas.

Six checks before putting real work files into Gemini

1. Test the exact file type and device

Do not infer that a PDF edited on a phone means a Google Sheet or presentation can be edited there too. Reproduce the same short task on web and mobile and document which controls actually appear. Google draws that platform distinction explicitly in its 29 October release.

2. Verify layout, tables and formulas after editing

For a PDF, compare page breaks, tables, images, citations and signature fields with the original. For spreadsheets, independently recompute key totals and test formulas in the destination application. For presentations, check fonts, alignment and slide order after export. These are recommended quality controls, not claims of known defects in Gemini.

3. Decide whether the document may be uploaded

Before uploading customer records, compensation files, contracts or employee documents, confirm the account type, organizational policy and applicable data-handling terms. Google's Gemini Apps Privacy Hub explains how consumer activity and uploaded material may be handled depending on settings. Do not assume the consumer Gemini app has the same contractual safeguards as a separately governed enterprise deployment.

4. Understand upload limits and supported inputs

Google's file-upload help page describes supported inputs and limits, including up to ten files per prompt subject to availability and a 100 MB limit for most non-video files. These upload rules are not proof that every file type can be edited with identical fidelity. Check the current help page and the product UI for the relevant account before designing a high-volume workflow.

5. Keep human approval for high-impact changes

A highlighted sentence may be easy to rewrite, but an AI-generated revision to a legal clause, employee policy, financial forecast or customer commitment still needs a responsible reviewer. Use a tracked before-and-after comparison, maintain an approved source copy, and require sign-off for material changes. Google's Canvas guidance documents editing and version navigation, but does not replace an organization's own document-control procedure.

6. Verify export, access and record retention

Test how the finished file is saved or exported, who can access it, and whether its final format meets the team's archive and accessibility requirements. Do not assume native-file round-tripping, electronic-signature validity, audit logging or enterprise retention policies merely because an in-chat preview is editable. Treat these as acceptance criteria for a pilot rather than features established by the product announcement.

Availability, pricing and what is still unknown

Confirmed: Google's 29 October 2026 post says the capability is available to all users in languages where the Gemini app is available; PDF preview and editing are supported on web and mobile; other listed formats are web-only. Not established by that post: a complete country-by-country feature matrix, organization-specific admin controls, identical limits across account types, native-file export fidelity, enterprise compliance certification for this specific workflow, or measurable time savings. It would be inaccurate to invent a special paid-tier requirement or claim universal enterprise readiness. See Google's release and the official privacy guidance.

A practical 20-minute pilot

Use three synthetic documents: a two-page PDF with a table, a short slide deck and a small spreadsheet with five formulas. On the web, ask Gemini to generate or open each, make one direct edit and one selected-text AI edit, then inspect the output. On mobile, repeat only the PDF scenario, because that is the supported mobile format Google explicitly identifies. Record whether each task succeeds, whether layout or calculations change, whether the final file can be exported in the needed form, and whether the account's privacy controls satisfy policy. A passing pilot is better evidence than a product demo.

Bottom line

For everyday work, the interesting shift is from AI that produces a block of text to an AI interface where a person can continue editing the artifact itself. That can shorten a workflow, but it does not eliminate review, source-of-truth ownership or information security. The sensible first move is to test one low-risk PDF and one web-based business file, not to upload an entire HR or finance document library. The original Google announcement is clear about what is newly supported and where.

Official sources

Google, 29 October 2026: generate, preview and edit files directly in Gemini; Google, 18 March 2025: earlier Canvas document-editing capability; Gemini Apps Help: creating and editing in Canvas; Gemini Apps Help: file uploads and limits; Gemini Apps Privacy Hub: data and activity controls.