Google on Thursday, October 29, 2026, gave users of its Gemini app a way to preview and edit the documents, PDFs, slide decks, spreadsheets and code that Gemini generates, without leaving the chat window. PDFs can be edited on the web and on mobile, while Google Docs, Google Sheets, slide decks, Markdown files, plain text files and code are limited to the web. The company said the capabilities are available to all users in every language the Gemini app supports, six months after it first allowed the app to produce downloadable files.

In Short

Google's Gemini chatbot can now show you the file it made for you, such as a PDF, a spreadsheet or a set of slides, right inside the chat, and you can change it there by typing, by highlighting a part and asking about it, or by asking Gemini to rewrite it. That matters to anyone who drafts work documents with Gemini, because the old routine of downloading a file, opening it in another program and returning to the chat to ask for changes gets shorter. On a phone only PDFs can be edited this way, and Google has not said how much editing free accounts get or which AI model does the work for them.

What Google published

The announcement is brief. It appeared on Google's Keyword blog under the Gemini App category, carries no author byline, and runs to four short paragraphs above a 39-second demonstration video. Its headline reads: "You can now generate, preview, and edit documents, PDFs, presentations, spreadsheets, and more directly in Gemini."

The opening line sets out the scope. "We're introducing new ways to create and refine documents, PDFs, presentations, spreadsheets, code, and more in the Gemini app," according to Google's post. There are no usage figures, no limits and no named product manager.

On the same day, the Google Workspace Updates blog published a companion note aimed at administrators of business and education accounts. That note adds most of the technical detail that the consumer post leaves out: what happens to each file type, which account types are covered, and how the rollout is staged.

How the editor works

Google describes three ways of working with a file once it appears in the conversation. "You can now generate, preview, and edit files right within your chat - expand to full screen to type edits directly, highlight text to ask questions, or have Gemini refine your files," according to the post.

The first is conventional: the file opens in a full-screen view and the user types changes directly, as in any word processor. The second turns a selection into a prompt. Highlighting a passage lets the user ask a question about that passage specifically, rather than about the document as a whole. The third hands the rewriting back to Gemini.

The study guide example

Google illustrates the feature with a single scenario. "As shown in the video below, you can now ask Gemini to create a PDF study guide from your uploaded handwritten lecture notes and then you can easily manually edit, highlight specific sections to have Gemini make edits to text or visuals, and even ask further questions," the post states.

Two details in that sentence deserve attention. Selection-based edits apply to "text or visuals," so a chart or image inside a generated PDF can be targeted, not only paragraphs. And the starting material is handwriting. Gemini runs on multimodal large language model technology, and reading handwritten pages from photographs is a harder relative of OCR, the long-established technique of converting images of printed text into machine-readable characters. Cursive, abbreviations and diagrams in lecture notes leave room for misreadings, which then flow into whatever document is produced from them.

Google's help documentation sets the boundaries for that input. According to the Gemini Apps Help centre, up to 10 files can be uploaded in a single prompt, each non-video file can be up to 100 MB, and rolling limits govern how much can be uploaded and analysed over a period of time.

Web versus mobile

The platform split is the most concrete restriction in the announcement. "PDF preview and editing is supported on both the web and mobile, while all other formats (Google Docs, Google Sheets, slide decks, Markdown files, TXT files, and code) are supported on the web only," according to Google.

Phone users therefore get a single editable format. Anyone who asks Gemini for a spreadsheet or a deck on a phone can still receive one, but changing it inside the chat requires a browser.

And what counts as "web"? Google does not define it. The company released a Gemini desktop program for Windows on September 10, and the Gemini app also runs as a macOS program. Neither October 29 post says whether the desktop programs get the full set of formats or the mobile subset.

Format by format

The Workspace Updates note breaks the feature down by file type, according to Google:

  • Slide decks: Gemini can produce outlines or slide content and show a preview of the layout. Users can edit the slides or request changes before opening them in Google Slides or downloading them as PPTX or PDF files.
  • Spreadsheets: Gemini can generate trackers, tables and data models, which open directly in Google Sheets.
  • Documents and PDFs: Gemini can draft PDFs, Google Docs, LaTeX, Markdown or TXT files, which can be edited inline on the web or in full screen, and all of them can be uploaded to Google Drive.

The PPTX option is new relative to April. When Google added file generation to the Gemini app on April 29, the 11 supported formats included Google Slides but not Microsoft PowerPoint's format. Slide decks generated in the chat can now leave Google's ecosystem as PowerPoint files.

From downloads to an editor

The October change builds on that April release, which Google attributed to Maryam Sanglaji, Group Product Manager for the Gemini app. The April feature produced PDF, Word, Excel, Google Docs, Google Sheets, Google Slides, CSV, LaTeX, plain text, rich text and Markdown files on request, and was available to all Gemini app users globally, free tier included. Files went straight to the device as downloads, and the Google formats could also be exported to Drive.

That design left a gap. A user who wanted a change had to download the file, inspect it in another program, return to the chat, describe the problem and wait for a fresh version. The October update moves the inspection and the correction into the conversation itself.

Google has also been adding review steps elsewhere. On October 19 it gave Gemini Notebook, the product formerly called NotebookLM, a co-creation flow that lets users change the outline of a slide deck or video before it is rendered. PowerPoint export was already available there, but only to top-tier subscribers: when NotebookLM moved to Gemini 3.5 in June, its 11 downloadable formats, PPTX among them, were limited to Google AI Ultra and Workspace business customers with the matching add-ons. The Gemini app's PPTX download, by contrast, comes with a feature Google describes as available to all users.

The relationship with Canvas, the side-by-side drafting and coding space Google added to the Gemini app in March 2025, is not addressed. Neither October 29 post mentions it.

Availability and administrator controls

Google's consumer post is short on eligibility. "These capabilities are now available to all users in all languages where Gemini app is available," it says.

The Workspace note is more specific. According to Google, the feature is available to all Google Workspace customers, Workspace Individual subscribers and users with personal Google accounts, and it is live immediately for both Rapid Release and Scheduled Release domains, the two update tracks Workspace administrators can choose between. Users of all ages who have access to the Gemini app can generate files.

There is no administrator control for the feature itself, according to the Workspace note. An organisation that does not want staff generating and editing files in Gemini can only switch the Gemini app on or off as a whole.

What the announcement leaves out

Several questions go unanswered in both posts.

Which model does the editing

Neither post names the model behind the editor. That matters because the Gemini app no longer gives every account the same models. Since October 9, accounts without a paid plan have been limited to a single model, Flash-Lite, under compute-based usage limits, while the AI Plus, AI Pro and AI Ultra tiers receive higher allowances. On October 13, Google brought Gemini 4 Argon and a tasks mode to the Gemini app for AI Pro subscribers, at $19.99 a month, and AI Ultra subscribers. "All users" may have the editor, but the same editing request could be handled by different models depending on the plan, and Google does not say whether editing requests count against the same allowances as ordinary prompts.

Uploaded files and Drive copies

Every editing example in the post concerns a file Gemini has just generated. Neither post says whether a PDF or spreadsheet uploaded from a device can be opened and edited in place, or only analysed and used as source material for a new file. Nor does either post say what happens after a document has been exported to Drive: whether later edits in the chat update that copy, create a new one, or stay inside the conversation.

Accuracy

The study guide example turns handwriting into a formatted reference document, a process with obvious room for error. Language models are prone to hallucination, producing confident statements that are not in their source material, and a polished PDF can make such errors harder to spot than a plain chat reply. The post does not say whether Gemini marks passages it was unsure how to read, or how users can trace a sentence in the output back to the notes it came from.

Limits

No file size limits, page limits or caps on the number of generated files appear in either post. The help documentation on file generation, which lists the April formats, describes no preview or editing step and carries no date.

Why this matters for marketers

Gemini's reach is the starting point. Alphabet said in its third-quarter results on October 28 that the Gemini app had passed one billion monthly active users, up from the 900 million it had reached in May. Comscore data published in September put Gemini at 30% of US AI prompt volume in June 2026, up from 17% in January, with ChatGPT falling from 70% to 50% over the same period. A feature switched on for that entire base, in every supported language, reaches more office workers than most standalone productivity tools.

The documents in question are the everyday output of marketing work: campaign briefs, media plans built in spreadsheets, client decks, reports. Agencies that send presentations to clients in PowerPoint can now take a deck generated in Gemini out as a PPTX file, and teams that keep budgets and trackers in Google Sheets can open Gemini's tables there directly. Gemini is competing on ground where OpenAI introduced a canvas editing space in ChatGPT in October 2024, Anthropic gave Claude the ability to create and edit spreadsheets, documents and presentations in September 2025, and Microsoft has built Copilot into Word, Excel and PowerPoint.

The editor also fits a broader pattern in which Google is turning the Gemini app from a place where questions are asked into a place where work is finished. Google added 14 more connected apps to Gemini on September 23, including Airtable, monday.com and PandaDoc, and on October 28 it opened a preview of its Connected Apps Platform to developers building for Search and Gemini, saying more than 50 connected apps were already live. Drafting, editing and exporting files inside the same window adds to the time users spend there.

None of that time is currently sold to advertisers. Google has repeatedly said the Gemini app carries no ads: after Adweek reported in December 2025 that Google had told advertising clients about a 2026 rollout, Dan Taylor, Google's vice president of global ads, said there were no ads in the Gemini app and no current plans to change that, and he repeated the position in January 2026. For now, features like this one feed Gemini's user numbers and its subscription tiers rather than ad inventory.

There is also a governance angle for agencies and brands that run on Google Workspace. Client data pasted or uploaded into Gemini can now become a finished, downloadable file inside the chat, and administrators have no setting that disables file generation and editing while leaving the rest of the Gemini app on. The choice, according to Google, is all or nothing.

Timeline

Summary

Who: Google, through its Gemini app team, with administrator details published by Google Workspace. The change applies to all Gemini app users, including Google Workspace customers, Workspace Individual subscribers and personal account holders.

What: Files that Gemini generates in a chat, including PDFs, Google Docs, Google Sheets, slide decks, Markdown and TXT files and code, can now be previewed and edited inside the conversation. Users can type changes in a full-screen view, highlight text to ask questions or request edits to text and visuals, or have Gemini refine the file. Slide decks can be opened in Google Slides or downloaded as PPTX or PDF files, and spreadsheets open in Google Sheets.

When: Google announced the feature on October 29, 2026, and said it was available immediately, including on both Workspace release tracks.

Where: In the Gemini app in all languages it supports. PDF preview and editing works on the web and on mobile; every other format is limited to the web.

Why: The update extends the file generation Google added in April 2026, which delivered finished files as downloads, by moving review and correction into the chat. Google does not say which model performs edits for free accounts, whether uploaded files can be edited, how edits relate to copies already saved in Drive, or what limits apply.